The 1994-95 academic year marked the second year of the FIPSE funding to create assignments, activities, and methods for using Perseus in post-secondary classrooms and to evaluate the effects that these creations have on students, instructors, and curricula. This year also marked the sixth year of evaluation of the Perseus Project and its effects as a technological innovation for teaching, learning, and research. Three main classes of results are reported here: the creation of a wide array of instructional resources based on Perseus (materials and methods), the fundamental challenges of creating and supporting physical infrastructure, and how Perseus changes the culture of teaching and learning. Although each of these results is addressed in this report, the main focus is on how Perseus changes teaching and learning and is fostering systemic change in different classics departments.
The use of Perseus in the 1994-95 academic year reinforced findings from previous years' findings and yielded indications of systemic changes due to the Perseus Project. The evaluation effort focused on five sites: Ball State University, Holy Cross College, Tufts University, Wesleyan University, and the University of Wisconsin. At each of these sites, Perseus was used in one or more classes, in some cases by more than one instructor. Data were collected in several ways, including:
* interviews with instructors and students (audio-taped),
* written journals kept by instructors,
* observations in some classes or labs,
* written questionnaires completed by students, and
* presentations by instructors at the Collaborative Meeting held April 28-30 at
Tufts University.
This report considers each site in turn, drawing descriptions of what was done at each site from the Perseus 1994-5 Annual Report.
One of the most tangible results of this year's work is the growing collection of materials and methods for using Perseus. Instructors created:
* new syllabi,
* assignments,
* Perseus paths,
* examinations and study guides,
* demonstrations,
* handouts of Perseus-augmented lecture notes,
* electronic presentations that incorporate Perseus materials, and
* a host of course and campus-specific instruction sets for using Perseus.
Instructors also used Perseus in a variety of ways, including:
* as the basis for scripted lectures,
* as an on-demand library for classroom lectures and discussions,
* as a supplementary resource for students, and
* as a resource for student projects which found expression in a variety of
forms such as Perseus paths, classroom group presentations, videotape
presentations, and WWW home pages.
This array of materials and methods represents enormous effort on the part of instructors and students, and illustrates the many ways that an electronic corpus of content and tools can be applied to teaching and learning. These materials and descriptions of how they were used are available in the proceedings of the Collaborative Meeting held at Tufts University April 28-30 and in the Perseus WWW site (http://www.perseus.tufts.edu). The web site section on "Teaching with Perseus" includes four example help guides for students on how to use Perseus in specific courses and campuses, and the syllabi and class notes for 16 courses taught by 9 different instructors.
Although these resources gather a substantial set of materials and approaches, they represent the tip of the iceberg of Perseus-based materials and methods. Below the surface for these instructors there are many years of learning what Perseus includes, what it costs, and what it enables--elements of what we have been calling conceptual infrastructure. In addition, there is a much broader Perseus-using community than what is represented in this evaluation. The Perseus list includes dozens of instructors who are using Perseus in various ways that are not included here. Also, new focused projects like the NEH-funded consortium of 5 universities led by Auburn demonstrate the impact that the project is having beyond the confines of this specific evaluation effort. The Perseus iceberg is not drifting aimlessly, however, it is growing and showing a path toward the evolution of technology-augmented teaching and learning--Perseus is bringing systemic change to specific programs and to an academic field.
This year's results again highlight the difficulties of introducing computing technology to academic settings. Instructors and students continue to be frustrated by lack of equipment, hardware and software failures, network incompatibilities, slow system response, and complex interfaces. The challenges of building and maintaining physical infrastructure were exacerbated this year by the upgrade on most campuses to Perseus 2.0 beta test materials. In addition to the now recognized difficulties of supporting Perseus use in classrooms, laboratories, libraries, and homes, the new challenge of upgrades were apparent in all the sites. Thus, the significant challenges of physical infrastructure that has dominated many of the previous Perseus evaluation reports was again a strong effect in 1994-5.
Culture of Teaching and Learning
In past years we have mainly reported how Perseus changed instances of teaching and learning. In this report we continue to document how Perseus enables instructors to incorporate more and better examples of content in their courses and how students are sometimes empowered to make discoveries and interpretations otherwise impossible or unlikely without Perseus. Examples of the former type of effect from this year include the juxtapositions of literary and visual mythical representations at Ball State, the application of Perseus' linguistic tools in Greek language at Holy Cross, the incorporation of geographical context at Wisconsin and Holy Cross, and the use of Perseus as a source of material for new presentations (e.g., Ball State) and study guides (e.g., Wesleyan & Wisconsin). Examples of the latter effects are best represented in the cases detailing student work at Holy Cross and the Web pages at Tufts. Thus, this year's results extend and reinforce how Perseus adds new potentials for changing the teaching and learning of individuals.
The longitudinal nature of the evaluation also provides a basis for assessing more global effects. In particular, we assess how Perseus changes the culture of teaching and learning. One clear indicator of how Perseus changes the culture of teaching is the ways instructors have begun to use Perseus as a basic resource upon which to build rather than as an add-on or centerpiece of a course--in essence, instructors are beginning to adapt Perseus rather than adopt it. Crane & Smith's development of Web-based courses that incorporate Perseus as one possible electronic source, Moskalew's incorporation of Perseus images into presentations for Roman poetry, Smith's development of geographic databases to extend the atlas features of Perseus, Martin's aim to make Perseus the "furniture" of the classroom, and Cahill's addition of specialized buttons to customize Perseus to his students' needs all illustrate a mature confidence to view Perseus not as a technological innovation but as a content resource available to meet their pedagogical goals.
Another indicator of changing teaching culture is how different instructors come to grips with the costs and benefits brought by Perseus (and all innovations). These recognitions are demonstrated by Martin's discussion of the "zero-sum game" of introducing Perseus at the expense of other course activities, Magrath's discussion of his concerted effort to more fully incorporate Perseus into a course over several semesters and the lack of improvement in student self-report evaluations, and Cahill's commentary about how much support students needed to use Perseus and their hesitation in trusting their own interpretations of primary evidence and dependence on secondary sources. As instructors gain more experience with Perseus they will continue to make more critical decisions about how and when Perseus is used and what objectives, materials, and methods will be weeded from their courses. Thus, we see a more mature culture of teaching emerging where Perseus is another powerful content resource as well as an adaptable tool for enhancing and changing the style of pedagogy.
Just as Perseus affects the culture of teaching as instructors gain more experience using it, there were some indicators that Perseus affects the culture of learning beyond the superficial behavioral changes of adding computers rather than books and lectures alone. Advanced students at Ball State discussed how Perseus use across several courses provided a sense of continuity and a resource for investigation on their own. They discussed the need for an introductory course for classics majors/minors that addresses the technical uses of Perseus so it can easily be used in subsequent courses and a Perseus-based capstone research course. These students have begun to recognize Perseus as a fundamental resource in a classics education rather than a system used in a single course. Just as instructors require multiple exposures to Perseus to begin to develop a sense of how it changes what is possible, we have evidence over multiple courses that students exhibit evolving expectations about Perseus. This evolution may be characterized by students moving from early states of awe and information overload to states of frustration and recognition of limitations, to later states of appreciation and integration. This evolution is evident in the comments of students at Ball State as well as by Magrath's discussion of student learning--beginning students are overloaded by all the information in Perseus and advanced students are frustrated that Perseus does not contain all the specific information they need. It is also explicit in Martin's discussion of a student at Holy Cross who was highly negative about Perseus in a 1993-94 course but blossomed in the 1994-5 course and asked to use her paper from the earlier course as the best example of her work at the school. As students continue to encounter Perseus in multiple courses, they will develop more mature expectations and capabilities with Perseus and additional confidence in trusting their own interpretations of primary materials.
These results for teaching and learning over time are important indicators of systemic change in specific classics departments. They illustrate that instructors and students are thinking differently as a result of Perseus. These results follow from the more tangible indicators of systemic change such as multiple instructors and multiple courses (e.g., Ball State, Holy Cross, and Tufts), the expanding number of instructors and courses around the world that are using Perseus (Yale University Press reported at the Collaborative meeting that the Perseus 1.0 edition was almost sold out), the development of commercial products that incorporate or refer to Perseus (new textbooks with Perseus activity options, other electronic products that build on it), and the emergence of a training service business for Perseus. Taken together, these indicators strongly suggest that Perseus has already changed the culture of some classics departments and is beginning to affect the larger arena of classics itself.
Ball State University: Evolving Expectations and Systemic Change
Ball State University (BSU) is an important evaluation site because there are multiple professors not directly associated with the Perseus Project development who have integrated Perseus into their courses. Over the past four years, the classics department has developed lab support for students (Perseus is available in CD-ROM version in two labs, one of which has trained assistants to help Perseus users) and in the 1994-95 academic year, they acquired a secure classroom kiosk and ceiling-mounted high-resolution projector in a classroom for Perseus use. These steps illustrate how BSU addresses the challenges of building physical infrastructure. This report focuses on how BSU is addressing the challenge of building conceptual infrastructure. Taken together, the BSU case best illustrates how Perseus is contributing to systemic change.
For the 1994-95 year, Professor Walter Moskalew used Perseus in his CC 105: "Classical World" course and Professor William Magrath used it in his CC 205 and CC 305: "World Myths" courses. Both professors (as well as Prof. Chris Shea who has used Perseus in CC 305 in previous semesters but had no Perseus-relevant courses in 1994-5 due to campus administrative responsibilities) and their students were interviewed during a site visit in the spring of 1995 and Magrath and Moskalew participated in the Collaborative Meeting at Tufts in the spring semester.
Professor Moskalew has used Perseus in CC 105 each semester for the past three years. He has developed two assignments related to vase paintings in Perseus (See the Collaborative Meeting proceedings for these assignments) and over the years he has increased his use of Perseus to support lectures. In the past, he had to bring a computer to class. The new classroom kiosk allowed him this year to use Perseus on a daily basis during lectures and discussion. This year he began to also use Perseus objects as part of electronic presentations (using Persuasion) in other classes. During the site visit, he enthusiastically demonstrated how Perseus images were incorporated into presentations and discussed Perseus as a resource that he takes for granted in preparing for classes. At the Collaborative Meeting he demonstrated a Persuasion presentation on Roman poetry that included images imported from Perseus. Thus, Perseus is serving as a library that Prof. Moskalew uses opportunistically to enhance his teaching.
Students in two CC 105 classes were interviewed during this year's site visit (March 31-April, 1995) after they had completed the first assignment and were in the midst of working on assignment two. Students reported spending about two hours on the first assignment, having little difficulty gaining access to Perseus in the laboratory, and getting knowledgeable support from laboratory assistants. This contrasts with previous years when access was difficult and lab assistants were not Perseus-literate. It should be noted that students were using Perseus 1.0 in the labs and Prof. Moskalew was using Perseus 2.0beta on the kiosk in class.
Assignments
When asked whether they would like to have more
assignments like this, one student said, "Yeah, I kinda liked it. It's a
lot easier than going to the library I guess, like everything you need is
there, like on this next assignment you have to get pictures, and all you have
to do is use the computer instead of going to the library and looking up
different books, card catalogue, all that; it takes that out of the way."
In response to a question about how these assignments differed from other
assignments, students noted that having illustrations was atypical. Students
were much more vocal about the use of Perseus in the classroom (at the time of
the interview, they were just starting the second assignment but had
experienced Perseus daily in class).
Classroom Use
Students said that class discussions were different
since the instructor could "show you when something is on a map" and
"click around" to illustrate points. Another student noted "they
(professor) can tell you what it looks like or give you a description but if
you have a picture, you see it yourself."
Prof. Moskalew provides handouts of the notes he uses in his lectures and most students found this to be useful. He also projects notes and students also found this to be helpful. One student said: "It's a lot better than a lot of classes; professors will lecture for a whole hour and you really don't know the key important points that might be on a test or quiz and this (projected notes and handouts from them) is all black and white and you know if its on these you better write it down because its going to show up." The ability to project and print lecture notes has been made easier by today's technologies and students will surely come to expect such devices as a matter of course in all classes in the future.
One complaint students had related to overload. One said, "He gives us more than we need to know, and in so many different directions." Even though outlines of notes, including maps were handed out to students, some still felt overwhelmed by the amount of information presented during class. Another student said: "I think he should try to be more selective about what he picks to show us instead of trying to show us everything he possibly can about that one particular thing; and then he goes so fast that we're trying to figure out what--if you are even able to write at all. You get frustrated and you walk out of here upset because you don't have anything." This complaint reflects observations at other sites where students in lectures were overwhelmed by the fast pace of verbal commentary and projected images. We have described this within the context of conceptual infrastructure that students and teachers must develop in order to be able to learn and teach with new technologies (e.g., previous evaluation reports; Marchionini & Crane, 1994). Instructors must find ways to integrate their verbal presentations with visual supports and to monitor student responses in order to adjust pacing. Students must develop attentional skills that allow them to take advantage of different media in parallel.
To the question of what was the best thing about Perseus, the immediate response from multiple students in one class was the images. In the second class, images were also mentioned, but the first comment was the availability of massive amounts of information; an interesting contrast to the overload comments of other students. The response to the question of what was the worst thing about Perseus that was echoed by several students in the first class was "too much information." In a subsequent discussion with Prof. Moskalew, ways to prepare students for new kinds of learning were discussed, specifically, talking to them about the learning process itself to prepare them for the feelings of overload that inevitably develop.
CC 105 leverages Perseus in the classroom to broaden the range of examples and illustrations. The two assignments are designed to lead students through vase paintings to draw inferences about styles. Although these styles could be simply stated in class and illustrated with a few projected examples, the assignments give students personal access to images that would otherwise have been inaccessible. The classroom use illustrates how students and the instructor are learning to take advantage of more primary information--one aspect of conceptual infrastructure. For instructors, the challenge is less a problem of finding good examples but filtering and selecting from the rich array. For students, the challenge is managing attention to take advantage of multiple channels of information. In the context of classics at BSU, CC 105 classroom use of Perseus and student assignments set a foundation of experience for students who take more advanced classics courses and projects an image of classics as a forward-looking department to students from other departments who take CC 105 as a core course.
CC 305: Perseus for classics majors/minors. The novice's too much is the
expert's too little
Professor Magrath has used Perseus in four sections of CC 305 in the past two years, including a section team-taught with Professor Shea in 1993. He also used Perseus in CC 205. In CC 205 a spiraled approach was taken by starting with a demo, then giving assignments based on the demo, then progressing to student independent use. In the CC 305 course, the progression of assignments moves from demo, sample assignment, group assignment, and finally a group project. The group project involves students in producing video tapes of their Perseus presentations. In the CC 305 course in the spring of 1995, about one-third of the students had previous Perseus experience in former courses. Thus, at Ball State, there is a progression of courses (105, 205, 305) in which Perseus plays a role. This is the best example of how Perseus has begun to influence an entire curriculum.
One of Prof. Magrath's goals is to help students develop visual literacy for reading vases and other objects. He used mythical representations on vases and in texts as foils to exemplify how the same concepts are expressed in visual and textual forms. In an interview, he discussed his goal to help students understand that a vase painting is an interpretation and that the medium changes the myth. He described his experience over several semesters and noted that students are at first overwhelmed by all the information in Perseus and after a while they cannot find all the information they want. He noted that Perseus 2.0 will elongate the first phase and wondered how to plan for this. This change in students' perceptions and expectations as they gain more experience is reflected in multiple elements of the evaluation. Furthermore, we argue that phylogeny recapitulates ontogeny as the perceptions and expectations related to the BSU classics curriculum evolve over time.
Assignments
During an interview with students in the spring 1995 CC
305 course, students expressed both similar and contrasting thoughts about
Perseus as did the less experienced CC 105 students. In contrast, some
students in the CC 305 thought that it was inconvenient to "get the
answer" in Perseus. One noted: "Sometimes it is slow, I could run to
the library and find a book and look it up." Several others noted the time
investment necessary to get to the lab to do Perseus work. Since at the time
of the interview, these students had done more assignments this semester
(although not the group project) and many had used Perseus in previous courses,
they had a better sense of the current limitations of access (must travel to a
laboratory, time restrictions on lab hours, relatively slow access using a
CD-ROM, etc.). Thus, these students were beyond the novelty effects
experienced by new users and accepted the computer as a common tool of the
course, focusing on practical limitations in their oral and written
assessments.
An interview with a student who had taken CC 305 the previous year yielded some interesting insights into the assignments and how Perseus affects learning. He estimated about 15 hours of use and said: "It was really, as a journalism major, refreshing to find something that catered to non-mythology-tuned people; coming into this fresh, the subject matter was a little hard to digest, and I'm really a visual person, so having Perseus with that whole visual backup really aided in my understanding of the subject matter." He said that he did a project on Achilles although he started with a project on Jason but found Achilles interesting as he browsed through Perseus. In describing how Perseus is different from using a library, he noted that the time commitment was much greater going to the library than using Perseus and went on to say: "The kind of information that is stored on Perseus, you have to get from several different sources, you'd have to read, you know, go to one Homeric work and then maybe it would have a Greek translation; but then you'd have it all in front of you in Perseus, you'd have both translations; and say you want to find out about art, you'd have to find another book about artwork; but with this (Perseus) its all with one, a great combination of everything". He later added: "It's really beneficial to have; everything at a glance, at your fingertips; I think it's incredible the amount of information and visuals."
A student who was a classics minor and a lab assistant was interviewed to determine what kind of problems students had with Perseus. In describing how students do Perseus assignments, she noted: "They look at it as a chore--It's something I have to do--instead of looking at it and thinking `Oh this is really neat, I didn't know they looked like this then.' I think this has a lot to do with the blocking of the learning because they think `I don't what to do this, I have to play with this computer,' instead of thinking `Oh wow! I'm looking at something that's probably 2000 years old and here is an actual picture of it." This comment reflects nicely the distinctions between how professors who know and love their subject think about the assignments they produce and the way that students think about them. She also described the most common types of problems students have. Students have trouble getting into the program, and then once in, deciding where to go and what to look for. She noted that "there is a lot of assumed knowledge," for example that if you want to learn about Herakles, you look it up in Appollodorus. She also described the problems students have when searching for words that have variant spellings.
This student used Perseus for three courses at BSU, although it was only required in one of her classes. She described how she used Perseus in a course on cults (that did not use Perseus) to relate Demeter to the Egyptian god Isis. She said: "So I used it to further my knowledge of what I already know." This is a nice contrast to the students who do assignments as a chore and do not have the basic assumed knowledge to get started with systematic exploration or searching.
Perhaps most telling was the interview with some senior students who has used Perseus in several courses. There was sentiment for a capstone research course based on the Perseus corpus. One student suggested a short 2 credit course in using Perseus for Classics majors/minors or a special section of the CC 105 course for majors/minors. Since some of the students were planning on graduate studies in classics, they were asked whether they expected to use Perseus in the schools to which they will go. Although none had considered to make Perseus availability a condition of application, they each said they assumed that Perseus would be available in any graduate program. One man said he would purchase Perseus in any case for his graduate studies.
Thus, we see a progression of appreciation for Perseus--both the potentials and the limitations of use--as students have more experience actually using it. Students first do required assignments (e.g., in CC 105 or CC 305), awed by the images and amount of information and sometimes frustrated by access; then as a result of more substantial assignments, they learn the limitations of content and access; and finally, they develop a more mature view of Perseus as a resource that they can use independently in a variety of classics courses.
Classroom Use
When asked about how they take notes, most students
signaled that they did not. One said: "He unloads so much awesome
information; I absorb a lot more just sitting there and listening and letting
the parts that are familiar enough just fall into place in my own head rather
than writing it down." The issue of overload was similar in this class as
in CC 105. Several students noted that there was overload in class and in the
amount of work (readings as well as Perseus assignments). Citing the drama and
painting, one student said, "it all correlates; it all has to do with one
another but its kinda hard to do it all at the same time." Considerable
debate over how much overload occurred in class followed this statement.
Students reported "pictures" as the best thing about Perseus and "technology"
(slow response, limited access) as the worst thing.
The class interview made clear how much students enjoyed Prof. Magrath's lectures. Students noted that the instructor's explanations, anecdotes, and style were the highlight of the course. In some ways, sharing a popular teacher with a new technology may reflect poorly on the technology; it is certainly the case that many students took CC 305 as much for the professor as the content and increased emphasis on Perseus changed some of their expectations about what class will be like. One student from a previous semester wrote on the questionnaire "My expectations changed due more to the class and the professor, who makes mythology modern and enjoyable/alive. Perseus does help to make it more easily accessible." Just as students' expectations change as they develop understandings about what is and is not contained in Perseus, instructors and courses evolve as well. The CC 305 course has changed, asking whether it is "better or worse" assumes a context-free comparison. In reality, the expectations of the instructor and students must be related to previous versions. The point illustrates the problem of evaluating a moving target--the context varies as well as the object of interest.
Longitudinal Questionnaire Analysis
Prof. Magrath used the Perseus questionnaires in CC 305 in the summer of 1993, fall of 1993, summer of 1994, and spring of 1995. In addition, Prof. Shea provided questionnaires for her summer 1993 section of CC 305. An analysis of these questionnaires follows. Table 1 presents Spearman Correlations for the 122 questionnaires received from these five classes. The variables are grouped into three clusters: demographic responses (variable names are in plain text), interface responses (variable names are in bold text), and performance responses (variable names are underlined). See Appendix A for the full questionnaire used at BSU. As we found in previous correlational analyses, there are generally large correlations among the interface and performance variables but not among the demographic variables and any others. In the BSU case, the following variables showed nine or more (out of 16 possible pairings) strong correlations beyond the .05 level of statistical reliability: CONFID (confidence in finding information in Perseus), INFOAMT (amount of useful information found in Perseus), INFONEED (likelihood that information needed was in Perseus), LEARNEAS (ease of learning how to use Perseus), LOST (frequency of getting lost in Perseus), RECOVER (ease of recovering when lost), PCONTRIB (how much Perseus contributed to learning), SATIS (satisfaction with learning), USEASE (how easy Perseus was to use after learning), VALUE (value of Perseus use to learning). Clearly, the usability issues (ease of learning, ease of use, and frequency of getting lost) are more directly related to students' perceived performance than is their previous computer experience or the amount of time they spent using Perseus.
| COMPAPPS | .0612 N( 121) Sig .505 | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| COMPEXP | -.2210 N( 121) Sig .015 | -.3261 N( 122) Sig .000 | ||||||||
| CONFID | -.1764 N( 121) Sig .053 | -.1403 N( 122) Sig .123 | .2064 N( 122) Sig .023 | |||||||
| IMAGES | -.1113 N( 120) Sig .226 | -.1135 N( 121) Sig .215 | -.0166 N( 121) Sig .857 | .0415 N( 121) Sig .651 | ||||||
| INFOAMT | .0889 N( 121) Sig .332 | -.1085 N( 122) Sig .234 | .0923 N( 122) Sig .312 | .4157 N( 122) Sig .000 |
-.1019 N( 121) Sig .266 | |||||
| INFONEED | .0307 N( 121) Sig .738 | -.0166 N( 122) Sig .856 | .1177 N( 122) Sig .197 | .3241 N( 122) Sig .000 | -.1980 N( 121) Sig .029 |
.4576 N( 122) Sig .000 | ||||
| LEARNEAS | -.1497 N( 120) Sig .103 | -.2155 N( 121) Sig .018 | .0954 N( 121) Sig .298 | .4362 N( 121) Sig .000 | .1216 N( 120) Sig .186 | .2874 N(121) Sig.001 | .1817 N( 121) N( 121) | |||
| LOST | -.0478 N( 121) Sig .603 | -.2491 N( 122) Sig .006 | .1401 N( 122) Sig .124 | .4824 N(122) Sig .000 | .1397 N( 121) Sig .127 | .4694 N( 122) Sig .000 | .2643 N( 122) Sig.003 | .5815 N( 121) Sig .000 | ||
| PCONTRIB | -.1328 N( 120) Sig .148 | -.1888 N( 121) Sig .038 | .1524 N( 121) Sig .095 | .4280 N( 121) Sig .000 | .1301 N( 120) Sig .157 | .5334 N( 121) | .3461 N( 121) Sig .000 | .3442 N( 120) Sig .000 | .3781 N( 121) Sig .000 | |
| PFEAT | .0869 N( 121) Sig .343 | .2317 N( 122) Sig .010 | -.2558 N( 122) Sig .004 | -.1589 N( 122) Sig .080 | .0944 N( 121) Sig .303 |
-.1612 N( 122) Sig.076 | -.0689 N( 122) Sig .451 | -.1359 N( 121) Sig .137 | -.0535 N( 122) Sig .558 | -.1422 N( 121) Sig .120 |
| ASNDIFF | COMPAPPS | COMPEXP | CONFID | IMAGES | INFOAMT | INFONEED | LEARNEAS | LOST | PCONTRIB |
| PHOURS | .0845 N( 121) Sig .357 | -.0119 N( 122) Sig .897 | -.2354 N( 122) Sig .009 | .0012 N( 122) Sig .990 | .1536 N( 121) Sig .092 | -.0207 N( 122) Sig .821 | -.0318 N( 122) Sig .728 | .0404 N( 121) Sig .660 | .1033 N( 122) Sig .258 | .0603 N( 121) Sig .511 | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| PUSE | .0771 N( 121) Sig .400 | .1667 N( 122) Sig .066 | -.2652 N( 122) Sig .003 | -.1821 N( 122) Sig .045 | .0477 N( 121) Sig .603 |
-.0842 N( 122) Sig .357 | -.0887 N( 122) Sig .331 | -.1649 N( 121) Sig .071 | -.0681 N( 122) Sig .456 | -.1713 N( 121) Sig .060 | |
| RECOVER | -.1412 N( 121) Sig .122 | -.0821 N( 122) Sig .369 | .1298 N( 122) Sig .154 | .4255 N( 122) Sig .000 | .0315 N( 121) Sig .732 |
.4395 N( 122) Sig .000 | .2969 N( 122) Sig .001 | .5007 N( 121) Sig .000 | .5508 N( 122) Sig .000 | .3692 N( 121) Sig .000 | |
| SATIS | -.1543 N( 121) Sig .091 | -.1479 N( 122) Sig .104 | .1310 N( 122) Sig .150 | .4415 N( 122) Sig .000 | .0855 N( 121) Sig .351 |
.6015 N( 122) Sig .000 | .2746 N( 122) Sig .002 | .3872 N( 121) Sig .000 | .3873 N( 122) Sig .000 | .7127 N( 121) Sig .000 | |
| USEASE | -.1825 N( 121) Sig .045 | -.2215 N( 122) Sig .014 | .1292 N( 122) Sig .156 | .4732 N( 122) Sig .000 | .0866 N( 121) Sig .345 |
.3625 N( 122) Sig .000 | .2367 N( 122) Sig .009 | .6694 N( 121) Sig .000 | .6220 N( 122) Sig .000 | .3290 N( 121) Sig .000 | |
| VALUE | -.1054 N( 120) Sig .252 | -.1978 N( 121) Sig .030 | .1819 N( 121) Sig .046 | .3711 N( 121) Sig .000 | .0598 N( 120) Sig .516 |
.6548 N( 121) Sig .000 | .3615 N( 121) Sig .000 | .4033 N( 120) Sig .000 | .4470 N( 121) Sig .000 | .6368 N( 121) Sig .000 | |
| ASNDIFF | COMPAPPS | COMPEXP | CONFID | IMAGES | INFOAMT | INFONEED | LEARNEAS | LOST | PCONTRIB |
| PHOURS | .5198 N( 122) Sig .000 | |||
|---|---|---|---|---|
| PUSE | .4835 N( 122) Sig .000 | .6889 N( 122) Sig .000 | ||
| RECOVER | -.0108 N( 122) Sig .906 | .0013 N( 122) Sig .989 | -.0659 N( 122) Sig .471 | |
| SATIS | -.2394 N( 122) Sig .008 | -.0439 N( 122) Sig .631 | -.2673 N( 122) Sig .003 | .4842 N( 122) Sig .000 |
| PFEAT | PHOURS | PUSE | RECOVER |
| USEASE | -.1082 N( 122) Sig .235 | .0526 N( 122) Sig .565 | -.1014 N( 122) Sig .266 | .5075 N( 122) Sig .000 |
.4506 N( 122) Sig .000 | |
|---|---|---|---|---|---|---|
| VALUE | -.2217 N( 121) Sig .015 | -.0091 N( 121) Sig .921 | -.2555 N( 121) Sig .005 | .4816 N( 121) Sig .000 | .7487 N( 121) Sig .000 |
.4656 N( 121) Sig .000 |
| PFEAT | PHOURS | PUSE | RECOVER | SATIS | USEASE |
A closer look at CONFID illustrates that confidence is strongly (<.01 alpha level) and positively related to 11 other variables, but strongly and negatively related to PUSE (number of times Perseus was used) and negatively related to PFEAT (number of Perseus features used). Thus, students' confidence in finding information in Perseus is inversely related to how much they actually use Perseus. Although this may appear on the surface to be counter-intuitive finding, a closer examination suggests that as students begin to understand the volume and range of information contained in Perseus and the various tools for using the system, they report more "realistic" ratings on the performance variables. Similar results are found for SATIS and VALUE. This explanation also parallels Prof. Magrath's comments (in bold above) about students' changes in expectations as they gain experience with Perseus.
Changing expectations also applies to instructors as well. Prof. Magrath reported at the Collaborative Meeting that he systematically studied how students performed with Perseus over the four semesters as he extended his own use of Perseus in class and as a larger portion of the class entered with Perseus experience from the CC 105 or other classes. His findings showed that students' work and course evaluations were no better as Perseus was more carefully integrated into the course. The questionnaire data reinforces this result. One-way analyses of variance across his four classes and also for all five classes were run for the questionnaire variables and of the various performance variables, only VALUE showed statistically reliable differences. These differences favored the early classes over the more recent classes where Perseus was more carefully included. (See Table 2 for the results of all four CC 305 classes; Group 1 was Magrath's Spring 1993 section, Group 2 was Magrath's Fall 1993 section, Group 3 was Magrath's Spring 1994 section, Group 4 was Shea's Spring 1993 section, and Group 5 was Magrath's Spring 1995 section.)
On the surface, this is disappointing considering the considerable investment in time and money to make Perseus an integral part of the course. However, these results reinforce the theme of evolving expectations as the first classes (groups 1 and 4 in the Table) in Spring 1993 were team-taught, had minimal Perseus use in the classroom, and Perseus was new for the instructors as well as students--all combining to make a highly novel environment. As students used Perseus more heavily in the Fall 1993, there was still little classroom use (due to the difficulty of setting up and using the computer and projector), and access was often frustrating (the laboratory support had not yet been well-developed--24 of the 31 responses to what was the worst thing about Perseus mentioned difficulty of using the system and/or poor lab support). Seven of the 31 questionnaire responses noted that there was not enough information in Perseus, thus, many students were learning that Perseus in not infinite. The third iteration in the Fall of 1994 reflects a stabilization of questionnaire responses, likely due to the instructor's efforts to use Perseus in class and more careful spiraling of the assignments. The fourth iteration in Spring 1995 showed more stabilization as means for most variables moderated toward the grand mean for all sections.
The variability in how different students feel about their Perseus experience is reinforced in their disparate comments on the questionnaires. Compare the contrasting comments of different students in the Spring 1995 section. One student wrote: "It helped to challenge my problem solving abilities. I had to think instead of letting the computer think for me." Another wrote: "It let me learn what I wanted to learn and at my own speed." Still another wrote: "Perseus has allowed me to correlate pieces of information together in different ways. A good example of this is during the final projects my group got into a discussion with Bill about a vase showing Menolaus being forceful in taking Helen on one side and Helen reluctantly going with Paris on the opposite side. This led us to a discussion of how Paris could have been an escape for Helen from an abusive relationship with Menolaus." In contrast, other students were negative about their Perseus experience. One wrote: "Because I spent so much time working or trying to work with Perseus, I has little time to complete the other requirements of this class. Not to mention the work I had to do in other classes." Another student wrote: "It made me not want to learn because it is so aggravating and slow." In a similar vein, another wrote: "It frustrated me to the point that I no longer cared about learning."
As noted above, simple comparisons across classes are only one metric in triangulating on the moving target of technology adoption and integration. As Prof. Magrath said in an interview, the course has changed as he consciously worked to integrate assignments and include Perseus in lectures. Now having Perseus at home has allowed him even greater access and changed the way he thinks about teaching. At the Collaborative Meeting he expressed some disappointment that the questionnaire analysis did not show "improvements" but recognized that CC 305 continues to evolve. Content, themes, and techniques come into courses as instructors and disciplines grow--in most cases as add-ons to what is already done in a course. Since the time available for classes does not increase, these changes force other content, themes, and techniques to be dropped. In CC 305 and BSU classics in general, Perseus as a content resource and as a technological tool (technique) has been added on and we are beginning to see integration as instructors and students struggle with what to leave out. Change is difficult and some of this difficulty is expressed in questionnaire analyses and instructor and student interviews. What seems certain is that there will be adjustments in the degree to which Perseus is used but that Perseus has become an integral part of the BSU classics curriculum. This is perhaps the most significant result of the evaluation to date.
That Perseus is integral to classics at BSU is reinforced by plans Prof. Shea is making for new courses built around Perseus. In an interview, she described an upper-level course on the Ancient City where students would come one day a week for individual appointments to discuss what other students already know about topics and design Perseus-based units to augment this knowledge. This plan envisions students and instructors as co-teachers to fill gaps where a standard lecture does not have enough or the right kind of information. Perseus enables this because it provides rich sources of primary material. She noted that "in archaeology, Perseus is better than our library--we will think of Perseus as the library for this class with the campus library as a supplement." One of the challenges to this and other new courses is better access for students and instructors--an ongoing physical infrastructure challenge.
Prof. Magrath in an interview noted that he has seen an increase in the amount of time he puts in with Perseus but noted that his satisfaction with Perseus has increased or remained steady. He noted that Perseus has had several effects on his courses. First, having it available in class has reduced some of the lecture time given to application of theory; "Perseus allows the course to become more topical," supporting the study of myth and cult, art, literature, and religion rather than only the theory of myth. Thus, having a broad range of materials supports a broader range of examples and illustrations. This reverberates with one of the goals of the Perseus Project to make more primary materials available to instructors and learners. Second, Perseus supports learning "how to read a vase as a conflation and compression of myth." He is able to illustrate how the medium changes the myth because having multiple examples from art and literature exemplifies how a vase painting presents an interpretation of a myth. Third, Perseus has led to more group work. Group work enables students to obtain ideas they would not have had alone, share their work with others, and forces them to be clear about their own ideas in order to express them to others. He noted that as a result, "you get brand new highs and brand new lows." Some students present great new insights and others banal insights; the banal insights are "teaching opportunities". He noted that "Every product (class project) has an element that takes me by surprise." The bottom line is that for Prof. Magrath, Prof. Shea, and Prof. Moskalew, Perseus has changed the culture of their courses and by extension their department. Regardless of the superficial costs of Perseus, these instructors and their courses have changed and it would likely be more difficult to go back to doing without Perseus than it was to get started using it.
In sum, BSU represents the most comprehensive portrait of Perseus use outside development sites. Over four years, we have seen iterations and evolution in specific courses and systemic change in the overall curriculum. The Perseus-enriched curriculum at BSU illustrates both the pain and pleasure of change as instructors strive to improve the quality of teaching and learning. The results so far clearly indicate some of the physical changes Perseus has brought and begin to indicate the more complex conceptual changes that innovation bring to courses and a curriculum.
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
3.1702 307.6495 310.8197 |
.7925 2.6295 |
.3014 . |
.8765 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
4.0435 3.8710 4.3333 4.1739 4.1429 |
1.7183 1.7077 1.5788 1.3022 1.7403 |
.3583 .3067 .3223 .2715 .3798 |
1.0000 1.0000 2.0000 2.0000 1.0000 |
7.0000 8.0000 8.0000 7.0000 8.0000 |
3.3004 TO 3.2446 TO 3.6666 TO 3.6108 TO 3.3507 TO |
4.7865 4.4973 5.0000 4.7370 4.9350 |
| Total | 122 | 4.0984 | 1.6027 | .1451 | 1.0000 | 8.0000 | 3.8111 TO | 4.3856 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
3.4245 81.6001 85.0246 |
.8561 .6974 |
1.2275 . |
.3030 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
2.4348 2.2581 2.1667 2.3043 1.9048 |
.6624 .9650 .7614 .8221 .8891 |
.1381 .1733 .1554 .1714 .1940 |
1.0000 1.0000 1.0000 1.0000 1.0000 |
4.0000 4.0000 4.0000 4.0000 4.0000 |
2.1484 TO 1.9041 TO 1.8452 TO 1.9488 TO 1.5001 TO |
2.7212 2.6120 2.4882 2.6599 2.3095 |
| Total | 122 | 2.2213 | .8383 | .0759 | 1.0000 | 4.0000 | 2.0711 TO | 2.3716 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
11.7698 144.6565 156.4262 |
2.9424 1.2364 |
2.3799 . |
.0556 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
1.9565 2.5484 2.7083 2.9130 2.4762 |
.8779 1.1787 .9991 1.1246 1.3274 |
.1831 .2117 .2039 .2345 .2897 |
1.0000 1.0000 1.0000 1.0000 1.0000 |
4.0000 5.0000 5.0000 5.0000 5.0000 |
1.5769 TO 2.1160 TO 2.2865 TO 2.4267 TO 1.8720 TO |
2.3362 2.9807 3.1302 3.3994 3.0804 |
| Total | 122 | 2.5246 | 1.1370 | .1029 | 1.0000 | 5.0000 | 2.3208 TO | 2.7284 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
17.8942 152.0812 169.9754 |
4.4735 1.2998 |
3.4416 . |
.0107 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
2.0000 2.6452 2.4167 3.2174 2.5238 |
.9535 1.1986 .9743 1.3128 1.2091 |
.1988 .2153 .1989 .2737 .2638 |
1.0000 1.0000 1.0000 1.0000 1.0000 |
4.0000 5.0000 5.0000 5.0000 5.0000 |
1.5877 TO 2.2055 TO 2.0053 TO 2.6497 TO 1.9734 TO |
2.4123 3.0848 2.8281 3.7851 3.0742 |
| Total | 122 | 2.5656 | 1.1852 | .1073 | 1.0000 | 5.0000 | 2.3531 TO | 2.7780 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
13.8404 139.7005 153.5410 |
3.4601 1.1940 |
2.8979 . |
.0250 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
2.2174 2.8387 2.7917 3.0870 2.2381 |
1.0426 1.1283 .9771 1.1246 1.1792 |
.2174 .2027 .1994 .2345 .2573 |
1.0000 1.0000 1.0000 1.0000 1.0000 |
5.0000 5.0000 5.0000 5.0000 5.0000 |
1.7665 TO 2.4248 TO 2.3791 TO 2.6006 TO 1.7013 TO |
2.6682 3.2526 3.2043 3.5733 2.7749 |
| Total | 122 | 2.6557 | 1.1265 | .1020 | 1.0000 | 5.0000 | 2.4538 TO | 2.8576 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 116 120 |
19.9331 147.6041 167.5372 |
4.9833 1.2724 |
3.9163 . |
.0051 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 20 |
2.9565 3.7742 3.0000 3.6087 2.7500 |
.8779 1.3835 1.1421 .9881 1.0699 |
.1831 .2485 .2331 .2060 .2392 |
1.0000 1.0000 1.0000 1.0000 1.0000 |
4.0000 8.0000 5.0000 5.0000 |
2.5769 TO 3.2667 TO 2.5177 TO 3.1814 TO 2.2493 TO |
3.3362 4.2817 3.4823 4.0360 3.2507 |
| Total | 121 | 3.2645 | 1.1816 | .1074 | 1.0000 | 8.0000 | 3.0518 TO | 3.4771 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
10.3165 101.5605 111.8770 |
2.5791 .8680 |
2.9712 . |
.0223 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
2.7826 3.3548 3.1667 3.6522 3.0000 |
.6713 .8386 1.0495 .9821 1.0954 |
.1400 .1506 .2142 .2048 .2390 |
2.0000 2.0000 1.0000 2.0000 1.0000 |
4.0000 5.0000 5.0000 5.0000 5.0000 |
2.4923 TO 3.0472 TO 2.7235 TO 3.2275 TO 2.5014 TO |
3.0729 3.6624 3.6098 4.0768 3.4986 |
| Total | 122 | 3.2049 | .9616 | .0871 | 1.0000 | 5.0000 | 3.0326 TO | 3.3773 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 116 120 |
7.5941 167.6455 175.2397 |
1.8985 1.4452 |
1.3137 . |
.2690 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 20 |
2.9565 3.4516 2.8333 3.3913 3.3000 |
1.1069 1.2868 1.1672 1.2336 1.1743 |
.2308 .2311 .2383 .2572 .2626 |
1.0000 1.0000 1.0000 1.0000 2.0000 |
5.0000 5.0000 5.0000 5.0000 5.0000 |
2.4778 TO 2.9796 TO 2.3405 TO 2.8579 TO 2.7504 TO |
3.4352 3.9236 3.3262 3.9247 3.8496 |
| Total | 121 | 3.1983 | 1.2084 | .1099 | 1.0000 | 5.0000 | 2.9808 TO | 3.4159 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
240.7009 3274.2253 3514.9262 |
60.1752 27.9848 |
2.1503 . |
.0789 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
8.6957 10.6774 7.9583 7.0435 10.3333 |
2.2650 9.4106 2.5619 3.6740 1.6833 |
.4723 1.6902 .5229 .7661 .3673 |
4.0000 4.0000 3.0000 2.0000 7.0000 |
13.0000 60.0000 13.0000 20.0000 14.0000 |
7.7162 TO 7.2256 TO 6.8765 TO 5.4547 TO 9.5671 TO |
9.6751 14.1293 9.0401 8.6322 11.0995 |
| Total | 122 | 9.0246 | 5.3897 | .4880 | 2.0000 | 60.0000 | 8.0585 TO | 9.9906 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
18880.4934 853297.9083 872178.4016 |
4720.1233 7293.1445 |
.6472 . |
.6300 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
9.6957 42.4516 12.0833 27.0435 24.8571 |
6.0335 142.4076 10.5085 103.1501 19.3398 |
1.2581 25.5772 2.1450 21.5083 4.2203 |
3.0000 4.0000 .0000 .0000 6.0000 |
25.0000 800.0000 40.0000 500.0000 90.0000 |
7.0866 TO -9.7839 TO 7.6460 TO -17.5620 TO 16.0538 TO |
12.3047 94.6871 16.5207 71.6489 33.6605 |
| Total | 122 | 24.3689 | 84.9004 | 7.6865 | .0000 | 800.0000 | 9.1513 TO | 39.5864 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
967.4670 8250.8690 9218.3361 |
241.8668 70.5202 |
3.4297 . |
.0109 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
7.3478 8.5806 5.7917 5.3913 13.6667 |
7.4138 8.2251 4.6060 4.8123 14.1680 |
1.5459 1.4773 .9402 1.0034 3.0917 |
2.0000 2.0000 2.0000 1.0000 3.0000 |
40.0000 30.0000 20.0000 20.0000 55.0000 |
4.1419 TO 5.5637 TO 3.8467 TO 3.3103 TO 7.2175 TO |
10.5538 11.5976 7.7366 7.4723 20.1159 |
| Total | 122 | 8.0738 | 8.7284 | .7902 | 1.0000 | 55.0000 | 6.5093 TO | 9.6382 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
19.0718 121.3217 140.3934 |
4.7679 1.0369 |
4.5981 . |
.0017 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
2.6087 3.1613 2.3750 3.4783 2.7143 |
.9409 1.0984 1.0555 .8980 1.0556 |
.1962 .1973 .2155 .1872 .2304 |
1.0000 1.0000 1.0000 2.0000 1.0000 |
5.0000 5.0000 5.0000 5.0000 4.0000 |
2.2018 TO 2.7584 TO 1.9293 TO 3.0900 TO 2.2338 TO |
3.0156 3.5642 2.8207 3.8666 3.1948 |
| Total | 122 | 2.8852 | 1.0772 | .0975 | 1.0000 | 5.0000 | 2.6922 TO | 3.0783 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
7.9203 150.1207 158.0410 |
1.9801 1.2831 |
1.5432 . |
.1942 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
2.4348 3.0323 2.7083 3.1739 2.8095 |
.7878 1.3288 .9991 1.2304 1.1670 |
.1643 .2387 .2039 .2566 .2547 |
1.0000 1.0000 1.0000 1.0000 1.0000 |
4.0000 5.0000 5.0000 5.0000 5.0000 |
2.0941 TO 2.5449 TO 2.2865 TO 2.6419 TO 2.2783 TO |
2.7754 3.5196 3.1302 3.7060 3.3407 |
| Total | 122 | 2.8443 | 1.1429 | .1035 | 1.0000 | 5.0000 | 2.6394 TO | 3.0491 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 117 121 |
11.9624 124.5049 136.4672 |
2.9906 1.0641 |
2.8103 . |
.0286 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 21 |
2.0000 2.7097 2.3750 2.9130 2.3333 |
.6030 1.1312 1.0959 1.0407 1.1547 |
.1257 .2032 .2237 .2170 .2520 |
1.0000 1.0000 1.0000 1.0000 1.0000 |
3.0000 5.0000 5.0000 5.0000 4.0000 |
1.7392 TO 2.2948 TO 1.9122 TO 2.4630 TO 1.8077 TO |
2.2608 3.1246 2.8378 3.3631 2.8589 |
| Total | 122 | 2.4836 | 1.0620 | .0961 | 1.0000 | 5.0000 | 2.2933 TO | 2.6740 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 116 120 |
14.6068 133.3601 147.9669 |
3.6517 1.1497 |
3.1763 . |
.0162 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 20 |
2.3913 3.0323 2.8750 3.4783 .1500 |
1.0331 1.2776 .7974 1.1229 .9881 |
.2154 .2295 .1628 .2341 .2209 |
1.0000 1.0000 1.0000 2.0000 2.0000 |
5.0000 5.0000 4.0000 5.0000 5.0000 |
1.9446 TO 2.5636 TO 2.5383 TO 2.9927 TO 2.6876 TO |
2.8380 3.5009 3.2117 3.9638 3.6124 |
| Total | 121 | 2.9835 | 1.1104 | .1009 | 1.0000 | 5.0000 | 2.7836 TO | 3.1833 |
| Source | D.F. | Sum of Sqaures | Mean Squares | F. Ratio | F. Prob. | |||||
| Between Groups Within Groups Total |
4 116 120 |
3.9984 117.7041 121.7025 |
.9996 1.0147 |
.9851 . |
.4186 . |
| Group | Count | Mean | Standard Deviation | Standard Error | Minimum | Maximum | 95 Pct Conf | Int for Mean |
| Grp 1 Grp 2 Grp 3 Grp 4 Grp 5 |
23 31 24 23 20 |
.8636 2.0645 2.1667 1.9565 1.6190 |
.8335 1.0307 1.1293 1.1069 .8646 |
.1777 .1851 .2305 .2308 .1887 |
1.0000 1.0000 1.0000 1.0000 1.0000 |
4.0000 4.0000 5.0000 4.0000 |
1.4941 TO 1.6864 TO 1.6898 TO 1.4778 TO 1.2255 TO |
2.2332 2.4426 2.6435 2.4352 2.0126 |
| Total | 121 | 1.9504 | 1.0071 | .0916 | 1.0000 | 5.0000 | 1.7691 TO | 2.1317 |
Perseus has brought changes to Holy Cross (HC) as Professor Thomas R. Martin used Perseus in a class for the second year and with the addition of Professor Neel Smith to the HC faculty. Prof. Martin used Perseus 2.0b during the Spring, 1995 semester in his advanced ancient Greek language course: "Herodotus", a seminar sized class (eight students) from the sophomore to the senior class. Professor Smith was engaged in developing the Perseus atlas and other geographic information system extensions for Perseus and cooperated with Prof. Martin in developing the physical architecture for Perseus at HC. Prof. Smith also built the web site at HC that is specific to Perseus (see http://perseus.holycross.edu/). The results presented here relate to Prof. Martin's systematic application of Perseus and are based mainly on the detailed journal he kept as part of the evaluation effort (see the Holy Cross web site for his journal from the spring of1994 and The Collaborative Meeting Proceedings for the 1995 journal), supplemented by two observational site visits, and student questionnaires.
For this course, Prof. Martin aimed to "make Perseus a part of the normal apparatus of instruction" rather than a special project in its own right. He and Prof. Smith discussed their goal of making Perseus "a part of the furniture" of the classroom. Prof. Martin wrote: "...my goal of having Perseus just be one of the tools (on the level of scholarly reference books, etc.) that a student is expected to be able to use to be a fully participating member of an advanced class." For the 1994-5 academic year, HC upgraded the classroom with room-darkening shades and rheostat-controlled track lighting. A new Macintosh was acquired to act as a server and the laboratory machines were connected to the Perseus server. One machine and a projector were wheeled to classes. This arrangements was an improvement over the previous year, but Prof. Martin notes that having access to Perseus in faculty offices remains a problem. Much of the discussion in his journal relates to technical problems (e.g., server crashes at critical class times) and some students noted problems with the system on their questionnaires. Thus, although HC moved forward significantly in developing its physical infrastructure, reliable and easy access in classrooms and labs remains a challenge.
The Herodotus course is an advanced translation course; an author course that serves as the "pinnacle of its curriculum in ancient Greek." Students had daily translation assignments and the assignments discussed here are those that used Perseus. Perseus was demonstrated in class and the first assignment led students to find all occurrences of the word "aitie" in Book 1. He reported that few students needed additional help doing the assignment and the class used the data collected to draw conclusions about specific connotations of the word. Although written concordances could have been used for this assignment, learning to use Perseus for this simple data collection task set the stage for later activities that have no manual analogs. Prof. Martin noted that "collecting data comes more easily than interpreting it" to reinforce the notion that Perseus is a tool to enable thinking but thinking and interpretation in and out of class are the actual goals of the assignment.
The followup assignment led students to choose their own words for investigation and present a brief presentation using Perseus to explain the most interesting citation found. Students were not able to make their presentations until the following week due to a server crash. Student presentations provide opportunities to discuss interpretation and "stress that they need to ask the next questions about the meaning of their data and not be satisfied with their first observations."
Prof. Martin demonstrated the morphological analysis tool, Greek word search, and Greek word frequency tool about midway through the semester and assigned students the task of creating a simple database (paper or electronic) for meaning and contexts for the different citations found for the word "doulos." Students were asked to use the word frequency tool as an aid in drawing conclusions about the word's meaning. He wrote: "I am impressed with the analysis of passages they have done. All members of the class are able to contribute something that they found in studying the instances assigned to them." As a followup, he demonstrates how the word frequency tool can be used to compare how Herodotus and Homer use the word.
The term projects were then assigned. Students were to investigate how Herodotus and another author use a word of their choice and compile a "top ten list of amazing customs or natural phenomena" with citations to Herodotus. For the first project, Perseus word lookup, analysis, and frequency tools were used. In the assignment statement, Prof. Martin addressed the data overload problem and requires students to limit their investigations judiciously and explain what choices they made regarding these limitations. This is one example of conceptual infrastructure--preparing students for a problem that the technology exacerbates.
In discussing the results of the course, Prof. Martin noted that "All the papers demonstrated that students could use the philological tools of Perseus 2.0 without apparent difficulty." He also reinforced the notion that simply finding data is not enough: "The new philological tools, especially the word frequency tool, attracted considerable attention. Only one student failed to grasp the essential point that the ratio of occurrences to total words in an author is the significant number, not the total number of occurrences; this student had been absent several times late in the term and therefore may have missed my stressing of this point in classroom demonstrations. A couple of students seemed so engrossed in the word frequency tool that in their papers they did not get much beyond making long lists of usages of terms in which they were interested for a large selection of authors. They seemed content to summarize the range of possibilities of meanings that they found without taking the next and necessary step of looking deeper into the evidence. "
Perseus Empowers Interpretation and Discovery
In his journal, he presents four cases of student work and describes what Perseus enabled students to do that was beyond what would have been possible with only paper resources.
"One student, however, did use this "grapeshot" approach to the authors in Perseus to good effect in studying the word idiotes (often translated as "private person"). By investigating the use of this word in all the genres in Perseus, the student felt that he could "propose a better definition (of the word)." I agree. He concluded that "whether the context is concerned with the military, the world of education, money, the courts, the realm of the royal, or the cults of heroes, a fitting translation for this word is 'not especially outstanding or extraordinary in any way.' " Thus, he discovered that idiotes indicates what I would call, in appropriating a term from linguistics, the unmarked category of the individual as opposed to the marked category. Without the tools of Perseus this student could not have made this discovery ."
"Another student was intrigued by what the philological tools showed did not exist. He began by investigated the term despotes ("master, ruler") and was intrigued to find it nearly missing from the historian Thucydides while not uncommon in Herodotus and Xenophon. Suspecting that Thucydides must use different terms to express the concept implied by this term, he used the English-Greek word search tool to find other such words. He was confronted by many possibilities, and in the end constraints of time limited the range of his investigation. His preliminary conclusion (which he rightly flagged as preliminary) was that Thucydides in fact seems to use a limited range of terms in this context, emphasizing the verb krateo. He concluded that he would need to "limit" his investigation further in order to take it further toward a meaningful interpretation and that his work left him with more questions than answers. Another way of saying the same thing is that a more refined hypothesis would be needed to go further on the topic and that true research has as one of its primary (and, I think, desirable) outcomes the generation of hoards of new questions."
"One student, who had taken my "Perseus class" last year, came in several times to discuss how to make sense of her data and what to look for at the next stage of her investigation (and never to ask how to use Perseus or complain about problems). Our conversations revealed that she had started to think about interpreting her data from a very early stage, constructing and refining what amounted to a metaphor to explain the pattern of the evidence she was finding (namely, that Herodotus presents the concept of catastrophe, symphore in Greek, as operating as if it were an infectious disease). Noticing the (in her description) "surprising" frequency of this term in Lysias, an orator, and wishing to compare two different literary genres (history and oratory), she studied the term in Lysias' works as well as in Herodotus. Her conclusion was that, although Lysias uses the term in a wider array of contexts than does Herodotus, a definite overlap occurs in the use of the term by both authors to signal the presence of a "crippling agent" in human affairs. The notion of agency is not linked, however, to a definite origin in these authors. In sum, I found the paper to be a thoughtful treatment of the evidence selected for review. The most significant result in this case for my money was that the student seemed very comfortable with this method of working. She was finding evidence, studying its meanings, contexts, and implications, constructing a hypothesis to test, searching for further material for meaningful comparison across genre boundaries, and drawing conclusions that were appropriately qualified. This seems to me to be the method of research. The value of Perseus 2.0beta in this case was to enable the student to use this method easily and quickly and, with its expanded array of philological tools, to inspire her to go further than would otherwise have been feasible."
"One other student produced a remarkable paper using this same basic methodology. She began by using the English-Greek word search tool to investigate laughter in Herodotus, a topic that we had discussed briefly a couple of times in class when it came up in our translation of the Greek text. Using the English-Greek word search tool, she identified six words in Herodotus that expressed the concept of "laughter." She hypothesized from preliminary data that Herodotus used these terms to imply mockery and disdain rather than (in her words) "what one would have generally guessed," namely, joy and humor and merriment. After analyzing all the instances in Herodotus, she concluded that none of them expresses joy and that the overwhelming majority of the instances explicitly or implicitly express mockery and disdain. She then turned to Homer to see if there was a precedent there for this usage. (She chose to look at Homer because in class we stressed the importance of Homer as a text informing the text of Herodotus.) She indeed found the precedent there, concluding that about half of the instances of laughter in Homer occur in this sort of context. Other instances in Homer do express merriment, however, she concluded. Thus she further concluded that Herodotus was both using Homer as a precedent for this usage but also making his text different in choosing to exclude the usage of laughter to express joy, etc. The remarkable aspect of this paper is that its conclusions about Herodotus and laughter essentially coincide with those published by a well-known classical scholar, namely, Donald Lateiner in his article entitled "No Laughing Matter," published in the professional journal Transactions of the American Philological Association for 1977. Since we on purpose did not read modern interpretations before doing these particular philological investigations, I am confident that this student had never read Professor Lateiner's article. In other words, she was able to gather data and then offer an interpretation of it that is validated by independent, professional scholarship. That an undergraduate could produce this result seems me to me very encouraging for the use of Perseus as a tool and an inspiration for learning through discovery. It is also encouraging that having data readily available in Perseus made it natural for a student to go on to do comparative work that went beyond the text of the one author with which the investigation began."
In these cases, we see examples of how Perseus easily generates data that is the fodder for analysis and interpretation; enables students to do scholarly research that generates new questions; helps students internalize the need for interpretation of data rather than the tedium of data collection; and make discoveries comparable to those presented by scholars in the field. These empowerments are reinforced in the statement one student wrote on the questionnaire. "It allowed me to research the way multiple authors use certain words and made their words easily accessible so I was able to do research using Perseus that would have been extremely time consuming without it. Without Perseus, then, I probably would not have done this research so I feel that it definitely helped my learning experience."
Prof. Martin qualified his commentary as follows: "Of course I am not saying that such work cannot necessarily be done without using Perseus. For the (relatively few) texts for which printed lemmatized concordances or lexika exist, such as Herodotus, one can study the instances of the occurrences of words using only the books. Yet no school will have multiple copies of such books so that a class of students could readily use them for research projects with maximum flexibility. The simple convenience of having texts and tools readily available in Perseus is no trivial advantage for students. Furthermore, the word frequency and English-Greek tools allow approaches to questions that no print tools can provide. Students may not realize that they are doing things with Perseus that no one could do before and come to take these possibilities for granted, but it still seems wonderful to those of us who remember only too well when none of these advantages had yet been created."
In addition to demonstrating how Perseus can dramatically augment what is possible in a classics course, Prof. Martin addressed several important issues in his journal. First, the many technical difficulties experienced are stoically summarized in the following reflection in a journal entry late in the course: "Our situation raises the general issue of whether institutions like mine are yet structurally prepared to use information technology in the classroom as "just part of the furniture." At this stage of the revolution, as it were, it looks like instructors in situations like mine are going to need to continue to do more themselves than I had hoped in purely practical maintenance of the hardware/software, if students are to have much success in using it." As at Ball State University, there is recognition of the level of effort that is necessary to integrate Perseus into courses, but a willingness to make the effort based on the early results of student work that starts with Perseus to press beyond what otherwise would have been possible.
Prof. Martin also made extensive comments on the challenge of adding Perseus to an already full course. He wrote about the "zero-sum game" faced by all instructors who add Perseus to a course. "The problem remains that it is difficult to allow enough time to read Greek in the traditional fashion of such courses and also spend enough time "modeling the behavior" for students in using Perseus. And to include Perseus has meant leaving aside other work that I would otherwise assign, e.g., bibliographic searches for modern studies using the available printed tools, especially L'Annee philologique and Classical Review. This is a zero-sum game; introducing new material into a course means spending less time on something else. I think this dilemma will only become more pressing as information technology provides us with more tools for our work." This reflection reinforces the discussion of systemic change in the BSU section above.
Just as at BSU, the Holy Cross classics program is undergoing systemic change due to Perseus. Prof. Martin noted the need for an "information technology in Classics" course in his department and made notes about how he will change the course in the future. He discussed how students evolve as they gain experience with Perseus, giving an example of a student from a previous course and how her expectations changed. "Just as it takes time for students to learn to approach technical and practical computing problems with equanimity and initiative, it apparently also takes a while for the value to sink in of the kind of work methods that use of Perseus promotes. This same student had written a thoughtful paper in last year's "Perseus class" despite her complaints, but she did not seem very pleased with it at the time. Some months later, however, she came to retrieve this paper to submit it as a writing sample for a fellowship application because, she told me, "it was the best paper I've done so far in college." Both these observations suggest to me that the optimum use of Perseus for major research projects will be in non-introductory courses, at least until students come to college already equipped with greater familiarity with computing technology than seems still the case. This level of expertise will be reached once students in secondary school are taught to use computing technology as a tool of research and not just as electronic wizardry."
In sum, at HC we see indications of systemic change reflected by changes in students during the semester and over multiple semesters; changes in how an instructor adjusts to the capabilities of a group of students and technical problems and in his expectations for future courses; and changes in how a department builds physical and conceptual infrastructure to support new approaches to learning and teaching. Most importantly, through the careful documentation Prof. Martin provided, we see examples of how Perseus empowers students to leverage the mechanical advantages that technology provide to create interesting interpretations for the content and discover insightful connections among the concepts central to the course.
Tufts is the development site for Perseus and thus Perseus has a significant but special influence on the classics department. Faculty and students work on the development of Perseus and thus courses offered by Prof. Crane and Prof. Merrill include Perseus in various prototype forms. Because this is the development site, most of the effort related to Perseus in the 1994-5 academic year went into developing the next iteration (Perseus 2.0) rather than into integrating Perseus into courses. Nonetheless, at Tufts this academic year, several classes made some use of Perseus: Professor Gregory Crane's Greek 103: "Greek Historians"; his new course, Classics 189: "Inventing Science: Greek Scientific Thought from Thales to Archimedes"; Professor William Merrill's Classics 37: "History of Ancient Greece"; Professor Susan Setnick's first semester Latin language course; Professor Betsey Halpern's Classics 75: "Classical Mythology"; and Professor Jodi Magness' Classics 164: "Greek Art and Archaeology".
Professor Crane's Greek course, a graduate level language course, used Perseus 2.0's morphological tools and lexical resources independently of any formal assignments. His course on ancient science, however, required students to use the World Wide Web both to do research with Perseus as one of the electronic resources and to present their reports and critiques of other students. In this class, Perseus was treated as one of the content resources that students could use in creating web-based expressions of their thinking (rather than traditional paper-based essays). The course syllabus and examples of student projects are accessible at http://www.perseus.tufts.edu/GreekScience/.
Professor Merrill's history course had an option of using Perseus for their final paper. A significant number of students were very interested in Perseus and used the Historical Overview both for their paper and for general review of the course material. This course yielded the largest number of Tufts questionnaires (15) and students reported use of Perseus ranged from 1 hour to 20 hours with a mean of 6.2 hours. Most students noted the images as the best thing about using Perseus and all reported good access in the Tufts laboratories. One student wrote "It makes it fun to be able to see places we're talking about. And if there's something unclear in class it's easy to use Perseus to clear it up." A correlational analysis of the questionnaires indicated that satisfaction and value were both strongly related (p < .05) to the perceived amount of information in Perseus.
Professor Setnick gave her Latin students an assignment in which they each had to research and then present orally their findings about a god or goddess of their choice. While they were not required to use Perseus, they were given a formal hands-on demonstration and then were offered personal tutoring sessions. About half the students actually used Perseus', tools, for instance the Encyclopedia and primary text references.
Professor Halpern's Mythology class had a similar assignment, in which they had to research and present orally anything (within limits) about classical mythology. They did not, however, have a hands-on introductory lesson, but instead an in-class demonstration. They were also offered, however, personal tutoring sessions and only one student took advantage of this. He was then overwhelmed with all the possibilities Perseus offered and the speed in which it did its searches. First, he used the English - Greek Word Search for "hero" and then examined the actual Greek definitions of each of those words to find the nuance he was looking for. Since he was pressed for time, he could not do as in depth a search as he had hoped for, but had not expected from Perseus.
Professor Magness' graduate level archaeology course also had the option to use Perseus for research and oral presentation of an object at the Museum of Fine Arts, Boston and hands-on introductory presentation. While they were also offered the personal tutoring sessions, no one took advantage of the offer. During the initial demonstration to them, many showed frustration that the database did not have "their" piece and did not seem to fully understand the depth of the database and the other ways in which they could have done their research.
In sum, the use of Perseus at Tufts is growing among the faculty as more of them understand what Perseus offers and encounter the "Perseus culture" that now sits in their midst. Systemic change at Tufts is thus stimulated by the presence of the project headquarters and the activity that surrounds a large-scale high-profile project.
Professor Carla Antonaccio had intentions of using Perseus in her introductory Greek language course in the fall, but Perseus was never fully and successfully installed on their Novell server. A site visit was then made during the January break by two of the Perseus staff in an effort to install and set up Perseus for the spring semester. Professor Antonaccio used Perseus in CCIV 201: "The Aegean Bronze Age". Again due to a combination of infrastructure problems, lack of computers for the 50+ students, and lack of relevant materials and working tools in Perseus 2.0, Prof. Antonaccio was not able to use Perseus to the extent she had hoped. She was able, however, to make Perseus paths for her students to use as a review for exams and quizzes.
Prof. Antonaccio maintained a journal (in the form of email notes to the evaluators) during the spring semester in which she described how Perseus was used and the technical challenges she and her students faced. After the network problems were solved, she was still frustrated by the lack of a projector. "One major frustration should be noted: my inability to bring Perseus into the classroom this term....I think that had I been able to use Perseus in front of the class, instead of having to rely on them to go to the TA sessions in the lab during the evenings we had space reserved, more students would have made more consistent use of Perseus." In spite of the technical constraints, limited amount of material relevant to the Bronze Age in Perseus, and the size of the class, Prof. Antonaccio was able to use Perseus in some novel ways. She and her TA set up paths that were posted on the network for students to use in studying for the quizzes and exams and she used the Perseus atlas to print maps for a quiz. These examples illustrate how Perseus can supply material that instructors use to support teaching--unlike the more typical use of Perseus to support lectures during class, this illustrates the use of Perseus to aid students' study and review and by the instructor to create assessment instruments.
Table 3 presents the descriptive statistics for the spring 1995 questionnaires.
Table 3. Wesleyan Descriptive Statistics
Valid
Variable Mean Std Dev Minimum Maximum N
ASNDIFF 2.46 1.05 1 4 13
COMPAPPS 4.62 1.20 2 6 16
COMPEXP 1.81 .83 1 4 16
COMPTYPE 2.25 1.34 1 5 16
CONFID 3.00 1.15 1 5 16
IMAGES 2.25 .93 1 4 16
INFOAMT 2.31 .95 1 4 16
INFONEED 2.44 .81 1 4 16
LEARNEAS 3.19 1.33 1 5 16
LOST 2.87 1.20 1 5 16
PCONTRIB 2.73 1.22 1 5 15
PFEAT 5.06 2.46 1 10 16
PHOURS 17.44 33.55 1 100 16
PUSE 7.38 10.23 1 30 16
RECOVER 2.88 1.20 1 5 16
SATIS 2.13 .96 1 4 16
USEASE 2.00 .85 1 3 15
VALUE 2.38 1.02 1 4 16
Professor Nick Cahill used Perseus in his Art History 301: "Greek Painting" course in the fall semester and Art History 115: "Greek Art and Society" in the spring. Professor Cahill gave detailed documentation to his students and made his teaching assistant and himself available for tutoring sessions. Prof. Cahill strove to teach his students that the process of finding answers is at least as, if not more, important than the actual answer. Perseus is a way to encourage people to concentrate on how, as well as what, they learn.
This Greek Painting course had 22 students. The goal of the course stated by Prof. Cahill in the syllabus, was to encourage the students to explore the methods of research: "The assignments are primarily looking rather than reading exercises, using primary materials - photographs and descriptions of the vases themselves - rather than secondary sources. You should look at the vases, read the descriptions, search the database, compare with other vases, but with the exception of the last assignment, you will need not do a lot of secondary research beyond the assigned readings, which will tend to be relevant to the project at hand." Four weekly assignments using Perseus were required of the class. Though they were intended to be group projects, Prof. Cahill explained "Only a few students actually worked in groups -- most I think did the work on their own, which is too bad. But people have such busy schedules..." The format of the assignments included the students following a path Professor Cahill had made that was relevant to the assignment's topic (vase shapes, symposia, mythology in Greek vase painting, and findspots of Greek vases) and then answering questions. The students had the option of handing in the answers on paper or in a path. None of the students, however, chose the path option. Due to lack of technology in the classroom, Perseus was never present during the class, but the computer lab was newly equipped with 65 Quadra 605 machines, all with good color monitors and ethernet connections.
In notes to an evaluator throughout the semester, Prof. Cahill observed the progress of the students with these assignments and reflected on the assignments themselves. After the first assignment, Prof. Cahill wrote: "The assignment had been written in terms of a series of questions to be answered, and that's what they did. The answers were fine and the assignments generally good, but I should have asked them to substantiate their claims more. In the next assignment I'm making it clear that I'm interested in the process of finding the answer, not in the answer itself -- that's why Perseus is different from reading a book. " These comments illustrate the subtle distinctions between assignment tasks and the presentation of assignments themselves. Reflecting on the second assignment, he noted: "This week we did an assignment on symposia. All responses so far have been written out -- nobody has attempted a path, which is OK with me. " After the third assignment, Prof. Cahill realized he should have structured the course a bit differently so that the students could get more out of the assignments: "I set the assignments up so that the students would do them *before* the relevant lecture. This was allegedly so they wouldn't be fed the 'answer' in class and regurgitate it later. I'm not so sure that this was such a good idea. My lecture on vases & mythology, for instance, tried to emphasize the independence of the artistic tradition and that you can't just consider the vases as illustrations of a monolithic Myth -- there were multiple versions of myths, and the vase painters had their own versions and agendas. Now I think that if I'd made this point before they did their assignments, they would have looked more carefully." For the final assignment, Prof. Cahill customized Perseus by adding a button to the Browser to save the list of sites in the Search Saver, so the students could then plot those sites on the Atlas. This exemplifies what instructors will do increasingly with open systems such as Perseus--adapt Perseus to specific assignments or adopt specific components of Perseus (e.g., BSU's Prof. Moskalew). Unfortunately, the beta version of Perseus was unable to support the assignment tasks in spite of the modification. Prof. Cahill wrote: "The big mistake here was in using a beta atlas which hadn't been tested. It's very frustrating for students who expect every site to be plottable, and also don't know whether something is a site or a region. So they click on Boeotia and don't know whether it's a site or what. Also the 'Plot sites in Search Saver' function of the atlas didn't work because it couldn't strip the '[14 vases]' off the site name; I fixed that too at the last minute, but too late for most students." However, Prof. Cahill was encouraged by that class: "We nevertheless had a good discussion of what was going on with trade and such. I think they're getting the hang of it, some of them at least. " This set of assignments and Prof. Cahill's written reflection on them nicely illustrate the rich interplay between teaching and learning--instructors learning from their teaching efforts and their students' learning with these efforts.
The course also required the students to read many secondary sources, some of which were only in Perseus and some of which were in Perseus as well as on paper. Professor Cahill spent an entire week making a path and a printed review for the midterm exam. For the review for the final examination, he gave the class the option of a path OR printed review and the students unanimously choose the printed version. The final paper did not require Perseus to be used, though it was strongly encouraged for both the primary and secondary resources it contains. In some of the student evaluations, however, it became clear that they realized how helpful Perseus can be in research. One student wrote on the questionnaire: "Perseus definitely helped me in writing my research paper. It opened so many doors to the many aspects that could be covered on my topic. To think of going through all the different books it would have taken to get the same in not even more info gives me a headache."
Independence or Lots of Hand Holding? (You can lead a horse to
water...)
Prof. Cahill was surprised how much hand holding of students was actually necessary for these assignments using Perseus. He distributed a detailed, illustrated, 8 page instruction manual describing Perseus from the moment the computer was turned on until the path was started [see Collaborative Meeting proceedings]. The notes within the paths were also very detailed and gave specific instructions on how to learn to use Perseus. After the first few weeks, however, he noted: "Again I've been alternately impressed by how much the students can do on their own, and despairing at how much help and hand-holding they need. Anyone using Perseus in a class had better be ready to lead the students through things very slowly, and to answer a lot of questions." He also observed that many of the problems the students were having with Perseus were due to the fact that just did not read or follow the instructions. This may also indicate, however, that the absence of Perseus in the classroom affected their retaining the instructions.
After the course had been completed, Prof. Cahill still felt that most students had failed to reach his goal as stated above. He reported to another evaluator, in reference to the final paper: "Some of them did quite well; I was surprised at how much they preferred to use secondary sources, though, and how much reluctance they had to looking at vases and using their eyes and thinking on their own." As the evaluators have observed since the earliest days of Perseus, students have different goals than instructors. Most students want to finish assignments and "acquire" knowledge. Since most of their experience in knowledge acquisition is in reading and listening to experts, the most direct way to achieve their goals is to apply those well-practiced strategies. Many students readily accept Perseus multimedia storage and lookup capabilities as valuable and adopt Perseus-based information-seeking strategies. What is more difficult and fundamental for students is basing thinking and knowledge acquisition on personal observation and inference rather than authoritative communications. In the Wisconsin case, we see students using Perseus to supplement authoritative readings and producing written essays rather than expressing their knowledge acquisition progress through Perseus paths. Some of these effects result from simply good time management, some are due to human resistance to change. The former is rational and addressable by educational innovators, the latter is more problematic and educational innovators seek methods other than forced practice to overcome knowledge acquisition skill inertia.
Technology: Computers bugs versus book time
There were a number of technical difficulties throughout the semester due both to the Novell server and the beta version of Perseus. Because of his experience as a Perseus developer, Prof. Cahill was able to handle most of these problems, as well as manipulate Perseus to fit his needs. A few of the students were also frustrated with the technology. Some of the things they found frustrating, however, were partly their own fault. For instance, on the evaluations, students complained of the locked gateway, Hypercard versioning problems, and certain features of Perseus that worked fine if you knew what to do, such as Enter Destination and the limiting keyword in the Browser. Prof. Cahill did hand out instructions and offered both his and his assistant's time for individual tutoring sessions. If these had been fully utilized, most of this frustration would have been avoided, as these were not real "bugs". Other things that frustrated the students were completely the responsibility of an imperfect beta, such as the Atlas and lack of on-line help, that have since been solved. The technological challenges are well-documented elsewhere but even with the technologically sophisticated Perseus developer, technical innovation is neither smooth nor automatic.
As is always the case, some students had negative reactions to computers, let along Perseus. Said one senior: "I think it could have [affected my learning] if I was more familiar with computers but as it is I spent so much time getting the program to work that I was too tired and frustrated to pay much attention to the actual assignments." Overall, however, the students were positive about Perseus and the opportunities it offered. Means on the five-point performance scales all clustered around 2 (1 is most positive, 5 most negative). Table 4 gives the descriptive data for these questionnaires.
Table 4. Wisconsin Descriptive Statistics
Valid
Variable Mean Std Dev Minimum Maximum N
ASNDIFF 2.18 1.19 1 4 17
COMPAPPS 3.50 1.76 1 6 18
COMPEXP 2.11 .83 1 4 18
COMPTYPE 1.39 .61 1 3 18
CONFID 2.44 .98 1 4 18
IMAGES 1.28 .46 1 2 18
INFOAMT 2.00 1.08 1 5 18
INFONEED 2.06 .73 1 3 18
LEARNEAS 3.17 .86 1 4 18
LOST 3.17 .79 2 4 18
PCONTRIB 2.11 .83 1 4 18
PFEAT 11.56 3.09 7 18 18
PHOURS 27.33 22.54 0 100 18
PUSE 12.22 3.62 6 20 18
RECOVER 3.28 1.13 1 5 18
SATIS 2.06 1.06 1 5 18
USEASE 2.39 1.14 1 5 18
VALUE 2.11 1.28 1 5 18
Perseus seemed to give the students at least the perception that they were able to spend more time actually learning. Consider the following responses to the questionnaire question "Did Perseus affect your learning? "YES! Able to see many examples of painters, potters, etc, without having to look through many books -- less time consuming, so able to look up more and learn more;"; "It [Perseus] made it easier to look at the vases in this class. By actually seeing them, different angles, etc, it was easier to remember"; "It [Perseus] shows that info can be consolidated for quicker and more thorough investigation." While the students may not have achieved Prof. Cahill's ultimate goal, they were able to detect that something was different, in a positive way, about this course as compared to others they've taken. Students wrote: "I learned to study from something other than books..."; "Perseus gave more insight into the class. I mean it made the class almost more hands on. You could actually see the image of what you were researching by the touch of a key"; "I really had to think about what I wanted to know and then figure out how to get that info out of Perseus." This last observation seems to be exactly what Prof. Cahill wanted these students to learn: sometimes the way you learn is more important than what you learn.
Greek Art and Society, Spring 1995
Professor Cahill's spring course, "Greek Art and Society", was a new, Perseus-based freshman seminar that attracted 11 students. As stated in the syllabus, Prof. Cahill's goals for the class were: "to learn more about Greek art, particularly vases but also sculpture, coins, architecture and other media; to investigate the roles such works of art played in ancient Greek society; to consider how scholars and students use works of art and other sources to learn about ancient society; to become familiar with scholarly research techniques, especially the use of computer databases in the humanities; and to gain experience presenting the results of your inquiries to an audience, and discussing the results of other people's presentations." Prof. Cahill's method of achieving this goal was: "to try to force them to work in groups and to look at vases and read ancient texts, rather than just look for articles in the library."
No previous knowledge of computers, the Classics or Art History was required, so there were wide varieties of backgrounds and class years, from freshmen Chemistry and theater majors to a senior majoring in marketing. Students were required to have email, however, and Prof. Cahill included a schedule of the Division of Information Technology classes for basic computer skills in the syllabus for students without experience. An email list was set up for the class also. The class explored a topic in Greek society, such as symposia or women, for one or two weeks. There were four groups and each was assigned a different aspect of that week's topic, such as women's clothing on vases and women in Xenophon's Economics. For an example of the division of assignments Prof. Cahill posted to the class email list, see the Collaborative Meeting proceedings.
Perseus was a part of every class, as groups were required to present their research in the form of a path to the rest of the class. After the second assignment, however, Prof. Cahill realized this was easier said than done: "The second [assignment] was to go through a vase shape and create a path explaining it. This was mostly to get people to create paths. What a disaster! Paths *really* need some work right now..." This again raises the problems of physical infrastructure, which is explored below. The class was held in a new high tech room in the Business School, which Prof. Cahill doubted would happen again easily, due to the rising demand for such classrooms. Because the emphasis was on research with Perseus instead of on content and therefore was unusual, the students were expected to put more energy into learning how to use and benefit from Perseus as a research tool than both traditional classes and most previous classes that have used Perseus. During site visit interviews, this was found to be a much preferred method of learning, as expressed by a computer science major: "It [Perseus] allows you to be creative and think on your own, I think...The book just explains why it has this on the vase or they draw this or why they put this, you know what I mean...I prefer Perseus [to traditional assignments]" Prof. Cahill both detected and shared in this enthusiasm after the second month of class: "Anyway people were pretty animated and interested. I'm looking forward to Monday because some people are going to come in with the impression that these [symposia] were polite dinner conversations where you talked about philosophy etc., and others are going to see them as wild drunken lust-crazed fraternity bashes. Then we get to sort out why different sources give us such wildly conflicting pictures."
Prof. Cahill's comments in an interview that "technology issues are very real" are driven home in detail in his explanations of how he works with Perseus. "I'm using a Duo, minidock, SCSI-ethernet connector, and a Luminator LCD projector. (It's a pain because in my office I have a big dock with ethernet card; in class I use the SCSI-to-ethernet, and at home I have a modem and SLIP connection, so I'm always reconfiguring the network setup). I'm mirroring the Duo's screen on the Luminator. But the mirroring doesn't come on immediately in the startup routine, and until the startup routine is finished, both screens are unpredictable: some things come up only on the Duo's screen, other things only on the Luminator's. What happened yesterday is that I got that error message saying that my appletalk wasn't working right -- because it was looking for a dock and a card and not finding it -- but the message was appearing on the Luminator screen, but not being displayed because mirroring wasn't turned on yet, so everything ground to a halt while it waited for me to press OK, but I didn't know to press OK. I thought I'd broken the dock. It was very painful. But this is the sort of thing that can stop a presentation dead in its tracks. Caveat professor! " He also listed other issues such as : getting a technologically suitable room, coordinating everything, having a good enough projector, all of which he says "puts extra burdens on the professor".
Prof. Cahill also made an important remark about conceptual interface challenges when he noted that as a Perseus developer, he already knew what kinds of questions Perseus could answer, but another professor with less experience in using Perseus would probably not know to ask these questions, since there had never been a way to answer them before. This makes more time commitment necessary for those professors. He continued: "[This] brings right to the fore this gap between ... power of resources and ability to take advantage of them." These comments reflect those at BSU and other sites as instructors grapple with how to teach with Perseus once they get it running.
Prof. Cahill also noted the problems with paths in Perseus: "Two groups reported that they made path stacks, and then would open them, only to watch their paths disappear -- I think they mean that the index card was blanked out. When they mailed the paths to me, indeed the index card was just blank. I don't know how that happens. So I fixed it by adding a script to the openstack handler of that stack, which checks that the index is in sync and if not rebuilds it. But then something else was screwed up, and I got a bunch of complaints that the paths didn't work properly. It's very frustrating for them, I think." However, Prof. Cahill stressed in an on-site interview that paths are very useful and should be made more flexible. For this reason, he added a function to paths which allowed his students to email their path which he could then easily add to the class' path stack.
Students also had concern about the stability of paths. One student explained this common complaint: "I try to do it all [the assignment] at the same time just because you would do it on computer and you're not always sure it's going to save." This same student one time had to give up and go home when she became so frustrated with repeated error messages. She did concede, however, that these problems were no "huge deal" when working with a partner. Most of the students seemed to accept that there were going to be problems that they might have to deal with. One explained: "I think some machines just don't work...The only problems are with sort of uncontrollable stuff like when your computer decided to not let you work or strange little errors come up."
Prof. Cahill, not expecting these students to continue on in the field, explained in an interview his expectations of the class: "How you break down a big topic like women in Greek society into questions you can answer. How are they depicted in vases? What is the relationship between the depiction of a woman on a vase and what the vase is used for? What does it tell about the audience ..(inaudible)..about women? How you use and read the primary sources...in a very critical manner. Perseus itself isn't going to tell you, but the absence of other secondary material...ought to be a spur for students to do this themselves...It's worked to some extent...It hasn't been something they take to instantly...they don't [seem to] trust their eyes". He was also, however, very impressed with the students and how they "took the problems in stride."
Most of the students agreed that they preferred Perseus based assignments to just reading in books. Said one: "I'm glad we're using it because it's more interesting than books." Another student said: "It's really something...it'll be in the future at some point for all classes, the way technology is." Some of the students also realized the work they are doing would be nearly impossible to do without Perseus: "Books and stuff wouldn't have it all in there..." followed by another "it could be 25 books in all..." A freshman, with no Art History or Classics background explained: "There's no way you could do it if it wasn't on computer because like, you know, he'll tell us Monday to come back Wednesday with something on women and religion and that's, you know, if you had to go to the library and do that you would never be able to come up with anything, I don't think." He went on to commend himself and his classmates for their achievement: "I guess its kind of impressive that we manage to do the work we do...there's people like me. I know I'm not the only person who has no experience and we manage to hold our own."