by

Eric B. Dent, Doctoral Professor

Deborah Wharff, Doctoral Student

Jacqueline Blackwell, Doctoral Student

This study examines an aspect of online education rarely examined, the interactions between professor and student that take place via email, outside the virtual, public classroom. Interaction and interactivity seem to be critical to the effectiveness of online learning. Consequently, the email exchanges may be some of the most critical for the effectiveness of student learning and retention. All of the student emails from four graduate courses, rated highly by course-ending evaluations, were coded and analyzed. Most of the emails from the students were of confirmation or affirmation of emails the professor had sent. The web-based course software was found not to be a pedagogical issue, particularly for students who had had prior experience with an online course. This study affirms the importance of developing a sense of community and of having a professor engage regularly and empathically with the students. We hypothesize that one significance of email exchanges is that they allow tacit knowledge to become explicit for students.

The number of university courses offered via the worldwide web continues to grow. The percentage of universities using "asynchronous internet-based courses" tripled from 1995 to 1998. By 1998, 78 percent of public four-year institutions offered such courses (NCES, 1999). These classes differ in significant ways from other distance courses that may have been offered as little as five years ago. Most web-based courses use a software application designed specifically for the purpose of conducting the course. Popular applications include WebCT and Blackboard. These applications provide for all of the online interaction that takes place during the course.

Significant research has been conducted on these "classrooms" - how "rich" they are, whether the professor adequately facilitates discussion, the frequency of interactions by the students, and so forth (Bowman, 2001). Interactivity and interaction are two important concepts in the distance learning domain. The use of interactive technologies allows learners and instructors to engage in learning experiences using high-performance computing and communications systems (Wagner, 1997) such as computers, cellular telephones, and the World Wide Web, thus enabling new instructional and learning experiences. These experiences include multimedia and hypermedia, computer-supported collaborative learning, interactive knowledge webs, and virtual communities (Dede, 1997) where interactivity emerges from the technological capability for establishing connections. Interaction, on the other hand, evolves from individual and group behaviors that directly influence one another, either point to point or point to group. These distinctions are critical because of the influence both interactivity and interaction have on the learning process.

This study extends that inquiry by analyzing an element of the course that is often not accessed by research. That is, an examination of the interactions that a student and professor have outside of the formal virtual classroom. The purpose of this research proposal is two-fold: (1) to determine the nature of the contacts that students in an online course make with a professor outside of "class." In other words, to understand the reasons why a student uses email to interact with a professor outside of the web course space. (2) to determine the nature of the interaction between a professor and student to see whether these exchanges support student success in the course. Interactions and interactivity enable the learner(s) and professor to develop relationships and build learning communities. Although these interactions have not been adequately studied, they may be vitally important. It is, perhaps, not an overstatement that these exchanges are more critical in terms of student retention and success in the course than the dialogue that occurs in the course.

One of the consequences of distance learning is the potential loss of the sense of community that is present in traditional classroom settings (Hiltz, 1998). Collaborative learning or active learning is a social process where students formulate new knowledge through communicating and interacting with others (Bouton and Garth, 1983; Alavi, 1994). Collaborative learning designs have proven to be more effective for online learning and have been correlated to enhanced student satisfaction with the learning and classroom experience. Recent studies provide evidence that collaborative learning is very important in creating positive performance outcomes for students (Hiltz, 1998). Positive experiences may influence students to repeat their use of distance learning as an instructional methodology. For groups to adapt to a structure of collaborative learning, professors play a critical role in modeling behaviors and encouraging interaction among students, and students must be willing to participate regularly in the collaborative learning process.

This study provides data about the nature of what may be critical dialogue between students and a professor. These emails, for example, include students contacting the professor because of a significant life event that is impinging on the student’s ability to complete coursework. The interaction about such events is analyzed and discussed. Secondly, this study generates important data for web course software design. The software used for the courses in this study is a proprietary University of Maryland University College product, WebTycho. The software includes the ability to support threaded conferences, streaming audio and video clips, student chat room, separate student study groups, assignment folders, and other typical features of this type of classroom software. Firstly, we conclude that, in general, the design of WebTycho works well as a learning platform for online education. Students have few issues with the software, especially after an initial learning period. Secondly, the issues raised center on the students’ inability to know that a professor has, in fact, received a particular assignment (although the software features a flag by which a professor can indicate to the student that she has read the assignment). A second significant issue is the "demand-pull" nature of WebTycho. As use of the web evolves, content most important to users is being "pushed" to them. As currently designed, WebTycho requires users to run the application to see whether there have been recent changes in the virtual classroom.

Finally, this study proves very helpful for pedagogical training for faculty using web course software. As with the issue of software design, this study finds that student emails do not provide insights into additional new pedagogical training. In fact, the primary meta-finding in this area may be that it is the customized use of email which provides the central ingredient for maintaining an effective learning community and addressing individual situations. The professor seems to have followed recommended practices such as rapid feedback, frequent interaction, personal empathy, and commitment (Kennedy, 2000).

Living at the beginning of web-based education is both a privilege and a challenge. The research challenge is trying to know anything about the experience when it varies so much. Little or no web-based education existed prior to 1993. In the mid 1990s, "web education" may have been essentially an experience with only email as the form of discourse. Even today, there are great disparities with some web education consisting of video files, PowerPoint presentations, sophisticated web pages that can lead students through exercises and others offering primarily a text-based interchange. The literature of the field is not detailed enough for the reader to ascertain the richness of the environment in order to make apples-to-apples comparisons.

The methodology of several studies to date is also not strong enough to merit definitive guidance in online education. For example, Phoha (1999) concludes that courses cannot be taught as effectively solely through email based on surveying 100 existing classroom students as to whether they thought the course could be taught as effectively via email.

A number of studies do hold the promise that the interaction between the student and professor is one of the most significant, if not the most significant variable in student success in an online course. In an empirical investigation that explored relationships between student perceptions and course design factors in 73 SUNY Learning Network courses, Swan (2001) found that interaction with instructors was one of the three factors which most significantly influenced students' satisfaction and perceived learning. Likewise, Gresh and Mrozowski (2000) identified interaction between instructors and learners as a critical element in the learning process. Her study focuses mostly, though, on the extra time it takes to teach online and warns that the necessary time for faculty-student interaction may get lost.

Another study of on-line learning suggests a missing link in the learning process that results from lack of dialogue and a sense of community normally derived from interactivity among students and a teacher (Roberson and Klotz, 2001). Effective use of technology, such as email, chat, and web based tools enhance interactivity and interaction in a virtual environment. Email provides a personal connection to and bond with the teacher and fellow students, thus creating a supportive environment for learning to take place. McLoughlin and Oliver (2002) identify three key elements of learning support in a distance learning setting that are required to encourage reflective thinking, social support for dialogue, interaction and flow of idea, and feedback from peers: social support, peer support and task support. Learning in a virtual environment can be assisted by different technological functions and should incorporate a variety of communication methodologies that support both process and product related outcomes to learning. A few of these outcomes are "just-in-time" support, collaboration and social dialogue, and multiple forms of learning activities (McLoughlin, 2002). The types of interactions to consider are: interaction to increase participation, interaction to build relationships and teamwork, interaction to establish expectations, interaction to exchange feedback, interaction to enhance elaboration and cognitive retention, interaction to increase motivation, interaction to negotiate and clarify understanding, and interaction for discovery and exploration (Wagner, 1997).

In a qualitative study of a graduate course in distance education at Arizona State University, McIsaac and Mahesh (1999) found that students displayed better commitment to their work in an online course compared with classroom courses because of the closer interaction with the professor online. Jiang and Ting (1999) collected survey data from students enrolled in 78 courses offered through the SUNY (State University of New York) Learning Network. Correlation and multiple regression analyses suggested that 11 independent variables contributed to students' perceived learning. Two of the variables were instructor-student interaction and instructor-student communication. A final portent offered of the possible significance of the interactions between the professor and the student comes not from an online environment but from courses offered by the Iowa Communications Network and via videotape. Webster and Miller (1997), using an eight-point Likert scale, found that most learners believed that experiences aimed at interaction between learner and instructor were slightly to very positively related to their learning. They also deemed instructors' interest, teaching skills, and personal interaction with learners important to the learning process.

Finally, it is important to note that several researchers assert the importance of the professor-student interaction primarily by focusing on its absence. Essex and Cagiltay (2001) found that even though some students found their online educational experience to be satisfactory, they experienced some degree of distress if there were insufficient online interactions with the instructor. Similarly, Dickie (1999) uses a phenomenological hermeneutic approach to describe her own experience as a distance learner. She makes ten recommendations of which one is, the human element must not get lost.

The sample of courses used for this study include four graduate courses taught by the first author at the University of Maryland University College (UMUC). Each of the courses was taught using UMUC’s proprietary software for web-based courses, WebTycho. The four courses and the semesters they were taught are provided below.

Spring 2001 - TMAN 633 HR Issues in a Technological Environment

Spring 2001 - OMBA 602 Individual and Group Dynamics

Summer 2001 - TMAN 633 HR Issues in a Technological Environment

Fall 2001 - OMBA 602 Individual and Group Dynamics

Email interactions occurred between the professor and student in the TMAN (Technology Management) courses (point to point). In the OMBA (Online MBA) courses, interaction took place between the professor and teams (point to group) and between the professor and the student (point to point).

The first author removed any identifying characteristics - name, employer listed, etc. - of the students before sharing the data with the other two authors. Each separate student was identified by a three letter code (not the student’s initials). The two OMBA sections were taught in a cohort MBA program. Because of the cohort feature of the MBA program, all students would have taken at least two online courses previously. TMAN students may or may not have taken a prior online course. An additional point of analysis is that the Summer semester course is compressed into a seven-week format compared with the other three which were offered in a fourteen-week format. For each of these courses, every email message sent to the professor by a student was retained. Likewise, every email message sent to a student was also retained. These archives provide a rich data set of interactions between a professor and student. It is estimated that the professor received only one telephone call per semester from an online student. Consequently, the email exchanges constitute nearly all of the dialogue that took place between student and professor outside of class.

Although this study does not examine course evaluation data, it should be noted that for all four courses, the professor received above average course evaluations. Work beyond this study could correlate the nature of professor comments with various elements of the course from the course-ending evaluation data. We make the assumption that the interactions of the professor reflect above average performance and are seen as generally positive.

Since this subject is relatively unexamined, we engaged in an exploratory content analysis (Weber, 1985) in which predetermined categories were not used (Krippendorff, 1980). Instead, consistent with a qualitative approach to content analysis, the researchers first immersed themselves in the data and then used their prior knowledge and experience to develop categories that emerged from the data. A two-level coding schema seemed most appropriate for the email data. This schema is provided below.

2. Query/ Request for Information

3. Withdrawal Notifications

4. Excuses

5. Searching/ Seeking (Non-Course related)

1. Deliverables (Assignment)

2. Assessment (Comments relate to course Activities Measurement of Learning)

3. Instructions/ Syllabus/ Textbooks (Comments general/ non-specific to #1 or #2 )

4. Instructor Feedback/ Response

5. WebTycho/ Library Online (Reference to media WebTycho accessible only)

6. * (Kept open for future use, as deemed necessary)

7. Learning/ Comprehension of Material/ Content

8. Email Usage

9. Extension

The latter two authors read each email to formulate tentative indicators for coding instances of the contextual markers associated with the two levels of categories. Incipient rules were recorded in a coding manual, which was periodically revised during the coding process as indicators and rules became more explicit. Two examples of coding manual guidelines are listed below.

The coding progressed in several rounds during which the researchers examined and coded sets of emails individually. After individual codings were completed, interrater agreement statistics were calculated and opportunities for increasing the agreement by creating additional coding manual guidelines were discussed. During these conversations, the coding rules were refined and articles about which there was seriously disagreement were recoded in light of refinements.

Cohen’s Kappa was used to calculate the interrater agreement (Fleiss, 1981; Shrout and Fleiss, 1979). In each case, the Kappa value was at least 0.72 and all statistics were highly significant. Fleiss (1981) considers Kappa levels above 0.75 as excellent. Given the exploratory nature of this research and the two-level categorization, these statistics are, at least, very good.

Note: The student numbers above are the number who finished the course. Some of the classes had emails from additional students who did not finish the course.

At level 1, over half of the emails from students were coded as affirmation/confirmation. The second largest percentage was for query/request for information. At the second level, most of the interactions were about the deliverables required for the course. The next two highest were instructor feedback/response (which doesn’t pertain to a specific assignment) and assessment (which does request feedback about ideas pertaining to an assignment). The numerical results for level 1 and 2 categories are listed in Tables 2 and 3, respectively.

Table 2. Coding results for Category Level I

There was a high rate of confirmation/affirmation statements. Because of the preponderance of these interactions, many examples will be provided below. These are responses to the professor and include statements such as:

Thank you for your patience

Your personal mission statement sounded incredibly interesting.

Your qualifications are impressive.

Thanks for all your kind words and efforts.

I truly enjoy the way you employ the conference section and involve yourself.

I appreciate your feedback.

I would like to welcome your baby into our world and church family.

I enjoy all of your feedback.

I think my tuition is really being well spent in this course.

Sir, thank you for the clarification.

Thank you, thank you, thank you. You have made me so happppy!

Thanks for sharing some of your life experiences and insights about PhD.

We appreciate your support in our continuing education.

Thanks for checking the availability of the McKnight article!

Sir, thank you so much for your expeditious response.

Thanks for the insight on the term cultural dynamics.

I really enjoyed this class and the concepts presented throughout. I also appreciate your help in applying to grad schools.

Thank you so much for everything it was a real pleasure, I learned a lot and am confident I will be a better person and manager/executive because of it.

Thank you very much for the way you conducted this class. I learned a lot and that will forever stay with me. I hope I will be able to call on you in the future.

After taking your class, I have realized that this is an area I would very much like to pursue further.

I like the amount of feedback you are giving our class. It’s great. Our last class we never had comments like yours from our professor. Thank you very much for the comments and discussion, it really does make a difference.

Thank you for your words of wisdom.

Also, thanks for taking the time to give me such a wealth of comments and ideas!

Thank you for your understanding and patience with me.

It is important for me to express how enriching this class was. Keep up the good work, you are in your right profession.

I will say that in all my years at UMUC (4 to be exact), you have been the most supportive and caring professor I have had. I thank God for you too.

I wasn’t kidding about learning more in these MBA classes than in my whole four year undergrad education combined.

Lastly, I would very much like to have lunch with you some time.

I have enjoyed your class and hope to have you as a professor in another seminar.

I look forward to meeting you sometime in the future. I would be interested to talk with you about future growth potential some day.

This subset of snippets from the student emails shows how appreciative they were of the responses they received from the professor. Nearly every student who interacted with the professor via email sent responses that were coded affirmation/confirmation. These snippets only show a single interaction. The following example provides the type of ongoing interaction that took place over several iterations.

Category 2 Coding	TMAN	OMBA
1. Deliverables	164	221
2. Assessment	41	39
3. Instructions/Syllabus/Textbooks	24	28
4. Instructor Feedback/Response	92	137
5. WebTycho media	13	24
6. Team related issues	0	32
Learning/comprehension of material/content	21	18
8. Email usage	2	5
9. Extension	13	2

Note: The September 11, 2001 attack occurred during the OMBA Fall class. Students exchanged emails with one another and with the professor to express their grief and sadness. Also, a deadly tornado hit the campus that same month and students exchanged emails on this subject.

Although they were not frequent in number, WebTycho issues are discussed here because one of the purposes of the research was to surface patterns in such issues. Following are the types of emails to the professor about WebTycho.

How to remove an entry in a conference area (2)

How to retrieve a power point presentation from a previous class

Please activate a study group

WebTycho menus were not accessible to student for submitting project assignment

WebTycho not accessible to student to post assignment

Pure voice software not downloading

Unable to post assignment to WebTycho

Request for video conferencing capability

Team study group established in WebTycho for posting assignments

Reserved reading article not accessible

How to post a bio in WebTycho

Affirmation that student was able to access WebTycho (3)

Not able to access library database article (3)

Query about where to post assignment

Unable to post assignment to WebTycho (2)

Could not locate mid-term exam in WebTycho

Unable to post mid-term exam in WebTycho

Change of email address in WebTycho

Since the OMBA students have had at least two prior web-based courses, whereas the TMAN students may or may not, it is not surprising that the emails pertaining to the fundamental use of WebTycho came from TMAN students rather than OMBA students. The latter were more likely to email about the software being down, the inability to post an assignment, or a reserved reading not available.

Throughout the database, the professor is seen as demonstrating empathy to the students on a variety of subjects unrelated to the course. Students contacted the professor to discuss topics such as: the professor’s personal website, the church, race relations, sick family members of the student, coping with a death in the student’s family, religion, keeping in touch after class ends, request for letter of recommendation, the professor’s motives in seeking a PhD in organizational behavior, and consultation on career decision. Although it is beyond the scope of this study, our research suggests that an important topic for future research is the relationship between the professor’s willingness to discuss with students these "off-topic" conversations and student retention.

Of notable interest, students in the OMBA courses discussed deliverables more than the students in the TMAN courses, but OMBA students requested fewer extensions. Working in groups might minimize the need for extensions, it would be interesting to do a study on this to determine if peer pressure to complete assignments on time leads to less requests for extensions.

The data collected and coded also provided a wealth of information to use as a means to evaluate and measure the number of different students engaged in email exchange with the professor, and the distributed frequency of email exchanges chronologically provided a foundational framework to assess interactivity in the distance learning domain.

Both recurrence (the number of emails from the same student) and variation (the number of emails from different students) were evaluated and assessed. In addition, the number of emails forwarded from the students to the professor was identified and counted. All student email communications received by the professor were acknowledged and rendered a response. Interaction was measured as a construct of variation, the number of email communications initiated by different students for the weekly time period intervals.

Per week periodicity was selected and used for data analysis purposes and for potential applicability across TMAN, OMBA, and other courses of disparate lengths. For each course evaluated, the first ‘student initiated’ email communication to the instructor, triggered the beginning of week number one. Equally, the last ‘student initiated’ email communication to the instructor, signaled the end of email communication for the course.

High interactivity and interaction periods resulted from the frequency distribution of the data sample collected, and may be valid indicators of the potential development or establishment of a ‘sense of community’ within in the distance learning environment.

For the TMAN 633 Summer course the highest state of interactivity occurred during weeks one, three, four, five, and six. For five of seven weeks total, high interactivity occurred as measured by the number of student emails exchanged. Of notable interest, 50 percent of the student population interacted with the professor via email communication external to the virtual classroom forum. Throughout the course, 28 of 30 students chose to contact the professor via email.

Table 4: TMAN 633 Summer student to professor email frequency and distribution.

Concurrent with high interactivity, high levels of interaction were also noted during these weeks.

Content analysis of the myriad of email communications exchanged during the high interactivity and interaction periods revealed an active and positive learning experience in progress. Email content exchanged reflected a substantial flow of ideas involving student perception, critical thought, and analytical discussion. For the compressed seven week TMAN 633 Summer course, the highest stage of both interactivity and interaction occurs in week three. The highest stages of interactivity, interaction, and ‘sense of community’ (both interactivity and interaction combined) occurs in a shorter cycle or at a faster development rate in the compressed 7 week course, as compared to the three 14 week TMAN 633 Spring, OMBA 602 Spring, and OMBA 602 Fall courses.

Higher stages of interactivity for the TMAN 633 Spring course happen in weeks three, five, ten, twelve, and thirteen. Higher interaction levels also mirror the weeks indicated with high interactivity. The highest state of interactivity occurs in the next to the last week of the TMAN 633 Spring course. In consecutive weeks, twelve and thirteen, the highest ‘sense of community’ is achieved. The ‘sense of community’ peaks in week thirteen, perhaps in alignment with final exams or course grades. In the TMAN 633 Spring course, email communications forwarded to the professor reflect a significant degree of active learning via the flow of ideas, new knowledge creation, and student personal responsibility. The frequency and distribution of emails from the TMAN 633 Spring course indicate a slower rate of evolving to a ‘sense of community’. Both interactivity (frequency) and interaction (variation) experience intermittent ebbs and flows.

The higher stages of interactivity and interaction for OMBA 602 Spring students occurs in week six, seven and thirteen, concurrent with the highest states of interaction. Of notable interest, for OMBA 602 Spring is that the highest sense of community is developed during midterm. More than 50 percent of the student population, interact with the professor at week seven. Similarly, TMAN Summer sense of community peaked for midterm weeks.

OMBA 602 Fall shows a few anomalies. Student communications began prior to course start and continue post course completion. A significant drop in interactivity is also reflected in week 12. The highest stages of interactivity and interaction occur in week two, three, and seven. More than 50 percent of the students interact with the professor in week three and week seven. The lower variation is consistent with the format for OMBA course work, in that all students are assigned to teams. Communications primarily occur via the student team group leader.

Oddly enough, the frequency rate and distribution variation by student emails did not occur at course start or at the onset of each major deliverable or exam for the courses evaluated.

The professor’s willingness to share personal aspects of his life resonates with some students, especially those who have family related issues and happenings in their life. For example, the professor’s wife was having a baby during two of the classes and the professor shared this information with the students. The professor and a student engaged in a dialogue around the role of families in society today. Likewise, several students provided feedback to the professor on personal matters, such as his church affiliation and academic pursuits or there own career frustrations or aspirations. These kinds of interactions seem to solidify the relationship between the professor and learner in order to create a "community" with valued bonds.

WebTycho is a passive pull system, meaning that students and professors must initiate actions to pull information from WebTycho on a periodic basis, (research, search for, and access new information which may or may not have been posted to the WebTycho system). This requires professors and students to check WebTycho frequently, constantly and consistently to determine if there are assignments, feedback, assessments, instructions, or course content discussions which might be relevant to the learner and professor. In the 21^st century technology age, when information is provided to the users through high-performance computing and communications systems with the support of intelligent agents, users rely on "push mechanisms" such as email and voice messaging systems to alert them of new and valuable "stuff" that relates directly to their informational needs and interests. Email and voice messaging communication modes continue to be the most used methods for gaining access to professors and other students in a virtual learning environment. WebTycho does have a feature that highlights with a red star those entries and topics that have not been read by the student or professor. This is one step towards helping the student and professor track interactions.

Any pedagogical training for faculty teaching web-based courses should begin with the "broad agreement among faculty and faculty unions that teaching online is more work than teaching in the classroom" (Kennedy, 2000). In The Chronicle of Higher Education Carr (2000) reported that "The key to having low attrition and successful completion in the online medium is the ability of instructors to keep the students engaged, and that requires quite a bit of effort from the instructor's point of view," he says. Successful distance-education professors e-mail their students frequently and respond to e-mail messages promptly, hold regular office hours -- whether online or in person -- and develop personal touches to make contact with their students, such as posting photographs of themselves on course Web pages. This leads to greater retention rate." This study affirms these findings. Effective web-based teaching requires a combination of a rich, shared learning environment with an empathic, rapid-response individualized interaction.

One possible explanation for the importance of email exchanges as a supplement to classroom experiences is that the email messages may help make tacit knowledge explicit. Tacit knowledge is described as "personal knowledge embedded in individual experience, involving intangible factors such as personal belief, perspective, and the value system" (Nonaka and Takeuchi, 1995). Formulated individually, personal, and context-specific; tacit knowledge is embedded in individual experience. Inherent association with intangible influences like personal beliefs, perspectives, opinions, and value systems, makes tacit knowledge particularly difficult to communicate to others in a standardized way. Context specific opportunity, motivation to share, and an external communication medium are prerequisites for exchange and tacit knowledge share.

Teese also recognized the tacit dimension of knowledge and associated the need of socialization or ‘sense of community’. (Teese, 1981, Polanyi, 1966) Student – professor interactions establish context-specific opportunities for idea expression and exchange during an online learning experience. Interactivity provides a channel or pathway for exchange. Interaction creates the medium through which tacit knowledge is converted into a transmittable, meaningful, and articulated communication flow of ideas. Teese also recognized the tacit dimension of knowledge and associated the need of socialization or ‘sense of community’ (Teese, 1981; Polanyi, 1966).

Web-based education is still in its seminal years. An important mindshift in teaching online is not trying to replicate what takes place in the classroom, but to try to create an equally or better learning environment online. We are only beginning to know what kinds of experiences and interactions constitute such a rich, fertile learning environment. At this point, it seems somewhat clear that the possibility of an ongoing dialogue with the professor that is empathic and responsive (in terms of needs, including time) is an important element of such environments for many students. The building of a "sense of community" is an element of the socialization process in action. Socialization influences student retention in the online learning environment. Equally, interactivity and multiple interactions as key indicators of the building of a "sense of community", also seem to significantly influence student retention in the online learning environment.

Weber, R. P. (1985) Basic content analysis. Beverly Hills: Sage Publications.