Abstract
Social presence (represented by salience of instructor and fellow students) in an online teaching environment has been shown to influence perceptions of course activity and self-estimates of learning. An experiment is described here in which teacher presence is manipulated via personalized messages from the teaching assistant throughout the course. High presence was not associated with activity in class discussion, homework performance, or tests over the assigned readings. It is suggested that these results have clear implications for the interpretation of research using perceived as opposed to objectively measured course outcomes to evaluate social presence and teacher immediacy concepts.
While a few well-funded universities such as Harvard and Stanford can offer Massive Open Online Courses featuring their top lecturers in live performance with professional camera work and editing, online instructors at most institutions do not have this capability. As a result, there is a problem with the notion of transferring the live lecture class to an online environment. Current studies indicate that students generally do not find the (low-tech) online lectures of sufficient interest, viewing these lectures does not improve their test performance, and students show little interest in even viewing the lectures after the initial weeks of the class (Grabe & Christopherson, 2008; Jensen, 2011). If lectures are not to be the main feature of many (if not most) online courses (as indicated by reference to the online instructor as a facilitator or “guide on the side” instead of the center of attention during instructional activities), then the question of teacher presence arises. Without the salient presence of the teacher, the online format degrades to an automated or “robot” course—with lists of reading assignments, online readings, short video clips, and quizzes with automated scoring and feedback. Yet in this writer’s experience, online course evaluations often include items that assume a high level of teacher presence (with items like “creates a feeling of online community” and “appears knowledgeable about the course subject”), and research indicates that students prefer their instructors to have a high presence online (e.g., Ke, 2010).
Research on teacher presence is closely associated with the more general concept of social presence in communication. Social presence theory aims to understand how electronically mediated communications could be effective in the absence of nonverbal cues (Short, Williams, & Christie, 1976). Social presence was described by Short as a construct based on intimacy (from eye contact, smiles, intimate topics) and immediacy (psychological distance between the communicator and recipient of the communication). As interest in social presence has picked up in the educational literature, it has been variously defined as “salience of the other person,” “being real,” “affective and psychological accessibility,” “affective connectedness,” and “ability to project oneself socially and emotionally into an [interpersonal activity]” (Kim, Kwon, & Cho, 2011). Theorists now must sort out the relevance of teacher–student and student–student relationships as components of the social presence construct, and the distinction between presence-related behaviors and presence-related perceptions must be made explicit. Until then, we are in a bit of a muddle as we confuse the sense of belonging (inclusion), feelings of attraction (motivation), liking for a course (satisfaction), and salience of other class members as we examine the meaning of social presence in a learning environment.
Student ratings of social presence in online courses have been reported to correlate positively with ratings of amount learned, satisfaction with the course, self-reported log-in frequency, and perceived interaction with other students and the instructor (Caspi & Blau, 2008; Kim et al., 2011; Richardson & Swan, 2003; Shin & Chan, 2004). All of the studies reviewed in this group relied upon course survey data. Picciano’s study represents the lone exception: At least one learning outcome was measured independently from the online questionnaire. Picciano created a measure of presence by combining questionnaire items addressing satisfaction, motivation, and feelings of inclusion. The resulting indicator of perceived presence was unrelated to performance on regular tests; however, it was positively correlated with grade on a class writing assignment (Picciano, 2002).
The previously mentioned studies addressed the broader social presence construct. Only a few studies have focused specifically on teacher presence (teacher immediacy) in an online teaching environment. Witt reported on a meta-analysis of 81 classroom studies in which teacher immediacy was related to learning outcomes. This analysis was apparently focused on traditional in-class settings (as opposed to online instruction). Teacher immediacy was separately defined as nonverbal (use of smiles, nods, relaxed body posture) and verbal (use of words that express liking or closeness). Teacher immediacy (verbal and nonverbal) correlated .49 and .51 with perceived learning but only .17 and .06 with direct measures of cognitive learning. The conclusion is that teacher immediacy influences student perceptions and attitudes regarding learning but has little actual effect on objectively determined learning outcomes (Witt, Wheeless, & Allen, 2004).
Of the studies reviewed here, none provided an experimental manipulation of teacher immediacy behaviors. Where learning outcomes were involved, they were almost always measured via student perceptions. Placing measures of both social presence (or teacher immediacy) and amount learned on the same survey instrument introduces the problem of method variance. To avoid such problems and contribute a causal contribution to the literature on teacher immediacy, the following experiment was conceived. The goal was to identify the effects of a set of specific teacher activities on objectively determined learning outcomes. It was hypothesized that introducing teacher behaviors indicative of teacher immediacy into a course will result in enhanced online discussion activity and higher test scores.
Method
Participants
The participants in this experiment were 132 students enrolled in an online critical thinking class offered by the Psychology Department at Humboldt State University in the fall of 2009. A total of 47 men and 85 women were enrolled. The class-level distribution was 48% first years, 26% sophomores, 11% juniors, and 14% seniors. This class satisfies a general education requirement for critical thinking skills, hence the high percentage of freshmen. Eighteen percent of the students were psychology majors, 75% represented a range of other majors, and 7% were undeclared.
Procedure
The course is taught fully online with assigned readings, weekly homework submissions, online discussion (organized into groups of 20–25 students and graded for quantity and quality of contributions), midterm exams, and a comprehensive final exam. With an initial enrollment of 150 students and only one instructor with one teaching assistant (TA) to handle the workload (weekly grading tasks and e-mail traffic), opportunities to provide individualized attention were limited. The decision was made to use part of the TA’s time to provide extended personal contact to half the class. This extra personalized attention was intended to combine two components of “teacher immediacy” or “teacher presence”—personal intimacy and instructor salience. The personalized attention suggested greater intimacy between instructor and student, whereas the frequent contacts made the instructor more salient as a feature of the course. (Note that the “instructor” in this manipulation was in fact the TA.)
To operationalize teacher immediacy, all students with last names beginning with A through K (N = 60) received the normal number of course-related messages from the instructor.
1
These included assignment reminders, brief feedback on homework submissions, prompts to stay involved on the discussion forums, explanations of grading, and general messages intended to motivate the students and keep the online course prominent on the students’ “to-do” lists. This constituted the low teacher immediacy condition. All students with last names beginning with L through Z (N = 70) made up the high teacher immediacy group. These students received the same level of basic attention provided by the A–K group, but they were also recipients of between four and six personalized messages from the TA. During Weeks 1 and 2, the TA uploaded a personal “reply” to each student’s personal introduction on the discussion forum. During Week 4, she sent each student a status check based on performance to date. For Week 6, she sent out personal reminders of her office hours and the availability of the professor and TA. Nonparticipants were sent a reminder of the importance of discussion board activity during Week 8. During Week 9, the less active students were sent a reminder to increase their participation. Finally, on Week 14, students with low exam scores were sent out some study tips and were reminded once again of available office hours. A typical “high teacher immediacy” message would look like the following: Hey Alissa—I just wanted to check in with you because I noticed you haven’t been checking your work against the example answer posted each week. It’s a good idea to do because it’ll help your understanding of course material (which will be helpful for the test), plus it’s an easy extra point =). Let me know if you have any questions.
Measurement of Student Performance
Three tests were given to measure understanding of the content from the two course textbooks: two midterm exams (multiple choice) and a comprehensive final exam (multiple choice plus a take-home problem set). Homework consisted of 14 weekly assignments that were graded on a 4-point scale emphasizing completeness of the submission.
The students were required to participate actively in the weekly online discussion of course-related issues. The total number of online contributions exceeded the time available for grading, so a sampling procedure was used where a random “ratings week” was selected from each 3-week segment of the course. This produced 5 weeks distributed over the semester in which every contribution to the group forums was evaluated and assigned a score. The TA evaluated the discussion posts using a 5-point scale that assessed depth of thought, adherence to the topic, responsiveness to the contributions of other students on the thread, inclusion of outside research, and length of submission. The total of these individual ratings constituted the student’s discussion grade.
Since it was possible for the TA’s ratings to be biased by knowledge of which experimental group a student was in, we added three objective measures of discussion performance: the total word count, mean number of words per post, and mean number of posts made over the five sampled weeks. (As it turned out, the TA’s ratings of individual discussion contributions correlated very highly with number of words in each posting: r = .96 based on a sample of 51 pairs.)
Results
The course grade records were examined for effects of the teacher immediacy manipulation. After removing students who enrolled but never actually got started with course assignments, we had an initial roster of 147 students and an end-of-semester roster of 132. In the high-immediacy group, 20 of the 76 students dropped the course, while 19 of the 83 in the low-immediacy group dropped. The difference in drop-out rates did not approach significance. There were no significant differences between the high- and low-immediacy groups on number of homework assignments completed or on scores associated with objective exams over the reading assignments (multiple-choice midterms and final exam). Examination of our records for the required online discussion revealed no significant differences in discussion ratings, total word count, mean number of words per post, or mean number posts made. And finally, we found no difference in extra credit work (participation in studies through the department’s research subject pool) between the two groups. (Access to course satisfaction ratings was not available due to the required promise of confidentiality.) The previously mentioned outcome measures were broken down by gender, and no performance by gender interactions were found with regard to our teacher immediacy manipulation.
Discussion
We manipulated the concept of teacher presence in an online critical thinking class by varying the number of personalized messages from the instructor’s assistant to individual students. It was hypothesized that a higher number of personal contacts would be associated with a heightened perception of teacher presence and enhanced course performance (as measured by discussion board activity and test scores over the assigned reading material). Such heightened teacher presence may motivate the students to become more active in online courses and learn more as a result. As it turned out, our course data did not support this hypothesis. Adding personalized messages to individual students (our manipulation of teacher immediacy) did not influence course activity, learning outcomes, or course drop-out rate. The lack of a teacher immediacy effect on test scores accords with the conclusion from the meta-analysis by Witt, Wheeless, and Allen (2004), where perceived teacher immediacy was found to be unrelated to cognitive test performance in face-to-face classes. Our finding on performance in online discussion is in disagreement with the research reported by Picciano (2002) and by Kim, Kwon, and Cho (2011), in which perceived social presence was positively correlated with grade on a writing assignment and with the amount of perceived course interaction, respectively.
One clear advantage of this study is that teacher presence was experimentally manipulated via specific instructor behaviors (actually TA behaviors) and the effects were observed on objectively determined criteria. This allowed us to avoid the problem of method variance associated with studies in which all measures are placed on one survey instrument. Almost without exception, the previous research on teacher presence consists of questionnaire surveys in which perceived social presence was related to perceived gains in learning and feelings of course satisfaction (with all measures on one questionnaire). It is tempting to conclude that course-related attitudes and perceptions are responsive to a different set of determinants than are “hard” criteria such as scores on tests of memory and understanding or quantitative measures of actual discussion participation. Cognitive learning may be more closely associated with time spent with the material and frequency of testing (Carpenter, 2012) than with social contacts (including teacher immediacy behaviors) within the context of the class. If the primary concern is with actual retention, understanding, and application of course content (as opposed to perceived retention, etc.), then future research should address specifically those instructor-controlled factors most likely responsible for the primary learning outcomes.
It is always possible that our failure to create enhanced academic performance is explained by a weak manipulation of teacher immediacy. Perhaps even more personalized messages should have been sent to the “high-immediacy” group—or maybe the instructor should have set up a schedule for Skype meetings with students in the experimental group. The control group (names A–K) received a number of reminder messages and suggestions for success in the course. Also, the instructor did respond promptly to all student queries received from individual students. This may have created too much of a teacher presence for our “low-immediacy” condition. The fact that the high-immediacy group (names L–Z) received a number of additional individualized communications with supportive comments may have simply been ineffective because a ceiling effect had already been achieved. But our intent was to use our normal level of teacher–student contact as our control condition and see if a personalized enhancement of contact would result in significant gains for our students using the resources (time and effort) available to a typical online instructor.
Those interested in following up on this study should include a manipulation check to verify that the teacher immediacy behaviors are reflected in a perception of enhanced teacher salience. For example, the course evaluation questionnaire could include a couple of items along the lines of “Rate the degree to which the teacher made his/her presence known in the conduct of this online class.” It would also be desirable to consider other ways of manipulating teacher immediacy besides personalized e-mail messages. Although recorded lectures are problematic (as described earlier in this article), several short web conference sessions, perhaps scheduled shortly before each midterm exam and the final, might be sufficient to make the instructor highly salient. These sessions could be recorded, capturing the chat activity as it occurs, so that students unable to participate could still profit from seeing and hearing the instructor at a time of their convenience. With respect to dependent variables, it is highly recommended that follow-up studies include both subjective measures (how much the students feel they have learned) and objective measures (word counts of forum posts, scores on multiple-choice exams). This will allow researchers to directly evaluate the issue of method variance.
Footnotes
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.