Home PublicationsJRTEPast IssuesVolume 33 Number 5

	Edited by Dr. David J. Ayersman, Mary Washington College, and Dr. W. Michael Reed, New York University
formerly Journal of Research on Computing in Education

Volume 33 Number 5 Summer 2001

Teacher Interaction
Motivating At-Risk Students in Web-Based High School Courses, Part I

Stephen Lehman, Douglas F. Kauffman, Mary Jane White, Christy A. Horn, Roger H. Bruning
University of Nebraska–Lincoln

Abstract
At-risk high school students working in a Web-based beginning composition course interacted for seven weeks with an online teacher in one of four different styles of interaction. Students received e-mails enhanced with motivation-building content and/or caring/personal-investment content. Students were observed as they worked in the course, and four variables—including ratings of electronic communication between teacher and student—were analyzed to determine student engagement in course content. Results indicate that enhancing motivation-building and personal-investment content in teacher communications increases at-risk student engagement in Web-based courses. Advantages of motivation-building enhancements over personal-investment enhancements and the importance of teacher interaction with at-risk students in Web-based environments are discussed.

Educators are increasingly turning to computer-based learning experiences to enhance instruction for their students. Proponents of technology typically cite such features as rapid access to information, increased capacity for authentic experiences, and enhanced opportunity to communicate. These potential benefits fit well with current sociocognitive views of learning. There is a growing worry, however, that an increasingly technologically based education system may not benefit all learners equally. Of particular concern are students who are at risk for academic failure for a variety of reasons, ranging from low socioeconomic status and learning disabilities to substance abuse and gang-related activities. The possibility of a widening gap—a “Digital Divide” between the technology “haves” and “have-nots”—appears to be a serious possibility (National Telecommunications and Information Administration, 1999), with limited access to technology standing as a barrier to student participation (Gladieux & Swail, 1999).

Arguments citing both advantages and disadvantages of technology-based approaches to teaching at-risk students can be readily constructed. Advantages include increased attention to relevant instructional material (Means, 1997), access to instruction outside the school setting for working students, and decreased potential for conflicted teacher–student relationships (Birch & Ladd, 1996). On the other hand, technology can make establishing productive teacher-student relationships very difficult. Though technology-based instruction seems likely to reduce the conflicts often present in teacher interactions with at-risk students, it would seem to have two distinct disadvantages.

1. Because of the technology, students are likely to have less access to the teacher’s scaffolding of unfamiliar material.
2. The barriers of communicating through technology have the potential to undermine the supportive relationships many at-risk students need in order to remain motivated.

The need for enhancing at-risk student motivation and monitoring of learning is particularly important because evidence seems to indicate that at-risk students’ self-initiated behavior often is not directed toward achieving academic success (Carroll, Durkin, Hattie, & Houghton, 1997; McWhirter, McWhirter, McWhirter, & McWhirter, 1993). In Web-based courses, which demand higher levels of self-regulation, at-risk students may fail to benefit from what Web-based instruction has to offer. These potential problems prompted the present study investigating the effects of teacher–student interaction on the motivation and performance of at-risk students.

Exploratory Study

In 1997, the Center for Instructional Innovation began its role as evaluator of Web-based courses being produced by CLASS, a Star Schools project of the U.S. Department of Education, targeting at-risk high school students. In CLASS courses, multimedia content is delivered across the World Wide Web as a part of the University of Nebraska’s Independent Study High School. Students submit assignments and interact with teachers by e-mail. In a preliminary evaluation of beta versions of these courses, we enrolled 16 students, half of whom were at-risk. Two of our observations regarding the courses are noteworthy.

1. There were large discrepancies in performance between the at-risk and not at-risk students. Many at-risk students disengaged despite support from on-site course facilitators, a stipend for working in the course, and forewarnings about the possible frustration of working in a technological environment.
2. The online teachers felt pressured as they interacted with multiple students in the online environment. The relatively large number of assignments from each student in some courses placed the teachers under considerable time constraints. In some instances, online teachers had little knowledge of or contact with the students beyond the exchange of assignments and grades.

Our strong impression was that at-risk students, in particular, needed more substantive interaction with the online teachers. In postsession interviews, at-risk students reported greater sensitivity to the limited contact with the online teacher than students not at risk. They also indicated that on-site facilitators were more important to their success than the online teacher. Their belief in the importance of on-site facilitators was presumably established through the higher level of contact with the facilitators. The importance of this personal interaction is generally consistent with literature in the field, which suggests that relational factors play a large role in the success rate of at-risk students (Brooks, 1994; McWhirter et al., 1993; Rak & Patterson, 1996; Werner, 1984).

Online teachers also felt frustration over the lack of teacher–student interaction. From their perspective, the high volume of graded assignments limited meaningful interaction with the students. Teachers described how they lacked a “feel” for students’ progress and expressed frustration in not knowing how to remedy this problem given their time constraints. When interactions with the teacher were minimal, at-risk students’ motivation seemed to wane and performance dropped considerably below their not-at-risk peers, whose motivation seemed less affected by their interactions with their online instructors.

In summary, the exploratory study indicated that student relationships with teachers and facilitators could have a substantial effect on the success rate of at-risk students in Web-based courses, but that the constraints faced by teachers of typical Web-based courses make developing these relationships difficult. This suggested to us that online teachers may need to take deliberate steps to optimize the interactions with at-risk students.

In the present study, we investigated the effects of variations in teacher–student interaction on at-risk students’ engagement in Web-based courses. Given the results of our exploratory study and based on the findings in the research literature (e.g., Goodenow, 1993; Kramer-Schlosser, 1992; Schunk & Swartz, 1993), we hypothesized that enhancing the level of meaningful interaction between the student and teacher would increase at-risk students’ levels of engagement in course content. We manipulated two dimensions of teacher–student interaction that we believed would enhance student motivation and engagement by adding: either motivation-building or personalized, caring interactions to the typical teacher-to-student communication. The efficacy of these interventions is borne out in the literature on motivating both at-risk and not-at-risk students in traditional classrooms.

Motivation Building

A great deal of research has addressed issues of motivation and student engagement in classroom settings, and much has focused on changing student beliefs regarding themselves and their academic performance. Self-efficacy theory and attribution theory represent a substantial proportion of this literature. These theories of motivation can be viewed in terms of Atkinson’s (1964) expectancy x value formula for motivation. According to this view, the motivation individuals display is a function of their expectancy of success in the task coupled with the value they place on the task. Self-efficacy and attribution theories focus primarily on increasing the expectancy factor of Atkinson’s formula. As students gain confidence that they can successfully complete the task at hand, their expectancy of success increases, as does their self-efficacy. Self-efficacy is influenced by the past performance, modeling, verbal persuasion, and the emotional state of the individual, and is often measured in terms of students’ confidence that they can perform a specific task. In attribution theory, expectancy is enhanced when students attribute the sources of their successes or failures to factors that they can personally influence. Attributions are rated as adaptive or maladaptive based on three dimensions of the attribution: locus of control, stability, and controllability (Weiner, 1986). One of the more productive attributions is attribution to effort, wherein a student attributes success or failure to the amount of effort that the student exerted on the task. In terms of the three dimensions of attributions, attributions to effort have an internal locus of control, can be changed (i.e., are unstable), and can be controlled. Students with high self-efficacy and/or adaptive attributions display increased effort, resilience, and persistence on educational tasks in traditional classrooms (Bandura, 1997; Pajares, 1996; Weiner). Given this evidence, it seemed likely to us that enhancing messages from teachers to students with efficacy and attributions-to-effort would also increase motivation and engagement for at-risk students in Web-based courses.

Caring/Personal Investment

The caring/personal-investment literature offers another perspective on the expectancy x value model of motivation. The students’ relational tie to the teacher may serve to enhance their value for the task because of their identification with the teacher (Noddings, 1992). Although less research has been conducted that frames the role of the teacher as a motivational construct, the research is growing (e.g., Battistich, Solomon, Watson, & Schaps, 1997; Midgeley, Feldlauffer, & Eccles, 1989; Moje, 1996, Noblit, 1993; Noddings; Wentzel, 1996, 1997, 1998). Findings in this area indicate that pedagogical caring enhances student persistence on academic tasks, and there is further evidence that this relationship may be particularly salient for supporting the resiliency of at-risk students (Brooks, 1994; McWhirter, et al., 1993; Rak & Patterson, 1996; Kramer-Schlosser, 1992; Werner, 1984). Results from our exploratory study also supported this hypothesis.

The importance of adult caring for at-risk students’ success is not surprising given the weight of evidence supporting the human need to form caring relationships. In a broad review of social and psychological literature, Baumeister and Leary (1995) found evidence supporting the hypothesis that the need for “belonging” (i.e., a caring relationship that extends over time) motivates human behavior to the extent that it constitutes a fundamental human motivation. Baumeister and Leary further argue that relationships are subject to satiation and deprivation effects, indicating that individuals with lower levels of relational connectedness may experience a greater need for such affiliative relationships. Thus, the motivational effects of caring may be greater for at-risk students than the general student population because at-risk students may experience fewer caring adult relationships than the average student (Birch & Ladd, 1996; Wentzel & Asher, 1995).

At-risk students’ need for caring adult relationships also seems to be borne out in the literature on at-risk resiliency. Werner (1984) notes that resilient students display protective factors that enable them to overcome the risk factors they face. In general, the most common of the protective factors discussed in the literature is the extent to which the student develops a positive, consistent relationship with at least one adult. Besides meeting the affiliative needs discussed by Baumeister and Leary (1995), these relationships may foster hope (agency and pathways) in individuals because of the perception they have of support and encouragement (Snyder, 1995; Snyder et al., 1991). This need may be more acute for at-risk students, whose success depends more heavily on this sense of support and belonging than that of students not at-risk (Kramer-Schlosser, 1992; McWhirter et al., 1993; Werner). Given this evidence, it seems probable that teacher–student interactions enhanced with personalized, caring content will increase at-risk student engagement in Web-based courses.

Our study investigated the utility of enhancing teacher–student interaction in online environments along two dimensions: motivation and personal investment/caring. We predicted that enhancing teacher–student communication with motivation-building statements and caring/personal-investment statements would increase student engagement in the course. We hypothesized that increasing motivation-building statements would increase students’ expectancy for success and subsequently increase persistence and time spent on task. We hypothesized that the teacher interacting in a caring/personally invested manner would enhance students’ sense of support, again increasing their expectancy that they could succeed at the course. Further, we hypothesized that the relationship developed between the teacher and student would increase the likelihood that the student would value the educational tasks the teacher and student worked on together because of the powerful effect of the need to belong and the students’ identification with the interests of the teacher. Thus, we posed three questions related to the two potentially motivating dimensions.

• Will enhancing teacher–student communication with caring/personal-investment interactions (e.g., focusing some interaction on developing a personal relationship with the student) influence at-risk students’ engagement in online environments?
• Will enhancing teacher–student communication with motivation-building interactions (e.g., pointing out past successes or attributing the student’s academic progress to effort) influence at-risk students’ engagement?
• Will there be combined effects where personal-investment/caring and motivation-building interactions become more effective when used together than if only one is used?

As part of a larger set of studies in which five Web-based courses were being field-tested, the present study used 16 at-risk students randomly assigned to one of four conditions in a course on beginning composition. The experiment took place in a naturalistic setting in which students were allowed approximately 50 hours (the estimated time to complete the course) to work in the courses and interact with the online teacher.

Method

Participants and Design

Sixteen at-risk high school students (13 female, 3 male) were assigned randomly to one of four conditions within a 2 (standard motivation versus enhanced motivation) x 2 (standard personal investment versus enhanced personal investment) design (Table 1). Participants in the four groups did not differ with respect to age (M = 16.06, SD = 1.18), grade (high school sophomores and juniors), or grade point average (M = 2.86, SD = 0.72). Participants were interviewed before entering the study to ensure that course content matched ability level. Participants included European American (n = 8), African American (n = 5), Native American (n = 2), and Hispanic (n = 1) students. All participants had previously been identified as at risk by a local social service agency, the public school system, or the legal system of a medium-sized Midwestern city. Participants’ risk factors included socioeconomic issues, teen parenthood, family issues (e.g., abuse and neglect), cultural identity issues, substance abuse issues, and gang affiliation. All participants reported minimal prior experience with computers and the Internet.

Table 1. Graphic Representation of the 2 x 2 Factorial Design with Experimental Condition Labels
		Personal Investment		Motivation Building
		Low		High
Low		Baseline professional		Motivating professional
High		Invested professional		Motivating, invested professional

Students received e-mail messages from the teacher containing either standard or enhanced levels of motivation-building statements and either standard or enhanced levels of personal-investment statements (Bosworth, 1995). High levels of each of these factors were considered enhancements to a baseline professional (standard motivation building, standard personal investment) level of interaction that we had observed to be typical of teacher–student communications in a Web-based environment.

Participants in the baseline professional condition received responses to questions and graded assignments without the benefit of motivation-building or personal-investment enhancements. The baseline professional condition represented a standard mode of communication typical of many Web-based interactions. An example of baseline professional communication is, “Yes, you should start writing here. If you have any questions about how to cut, copy, or paste, grab [an on-site facilitator].”

Participants in the motivating professional condition received messages containing self-efficacy and attribution-building enhancements intended to strengthen the students’ beliefs in their academic capabilities. An example of a motivating professional statement is, “The last three assignments you have handed in have been stellar. You are doing very well in this section.” In this example, the teacher used self-efficacy-building statements to try to influence the student’s engagement.

Participants in the invested professional condition received messages enhanced with statements indicating that the teacher was invested in the student personally, with a particular focus on relationship building. An example of an invested professional statement is, “If you can’t think of a topic, write about a movie you liked. Have you seen any good ones lately?” Here, the teacher focused on helping the student identify strategies for using her own experiences to guide learning. Additionally, the teacher expressed personal interest in the student with the implicit message that he valued that student’s life experiences and wanted to get to know that student better.

Finally, participants in the motivating, invested professional condition were sent messages that included both personal-investment and motivating enhancements. Participants in this condition received messages intended to build efficacy and foster adaptive attributions as well as demonstrate that the teacher was invested in developing a relationship with them. The following is an example of this type of interaction: “The way you put sentences together in that last message seemed very casual without being sloppy or lacking skill. Have you been writing a lot?” Later in the same message, the teacher related the assignment to the student’s involvement in sports. Here the teacher used efficacy-building statements in the context of a personalized interaction with the student.

Dependent Variables

Four variables were used to measure student engagement in course material:

• ratings of student effort on e-mails,
• lab monitor ratings of student engagement,
• hours worked in the course, and
• course grade.

The primary measure of engagement was student effort on e-mails. We measured effort because effort and persistence appear to be primary mechanisms by which motivation increases achievement. Rating the number, quality, and completeness of the e-mails created a basis for judging the effort students exerted in the course. We used e-mails because student e-mails constituted the entirety of the students’ interaction with the teacher and work on the course and, therefore, comprised the most direct measure of student course-related activity. We included lab monitor ratings to provide an account of student activities not communicated across e-mail. We calculated the number of hours worked to gain comparative information on the extent to which students persevered in course activities. Finally, we collected final grades as a global measure of the success each student experienced in the course. Using these related variables, including one based on direct observations of students working in the course, enabled us to triangulate our findings on student engagement.

E-mail Effort Ratings

Ratings of the level of engagement of student e-mails submitted to the online teacher comprised our primary data. We collected student-to-teacher e-mails during the last two weeks that students remained in their assigned condition, allowing the teacher approximately two weeks to establish a style of interaction with the student before data collection. We hypothesized that students who were more engaged in the course material would exhibit greater levels of engagement in the correspondence and assignments they sent to the teacher. The 165 student e-mails sent to the teacher during this period were rated on a 4-point scale by two independent raters blind to the experimental conditions. Raters judged the effort expended on the e-mail based on three criteria: length of the e-mail, idea development, and overall completeness of the assignment. Analysis of interrater reliability revealed acceptably high reliability ratings between the raters (r = .79). E-mail ratings were aggregated by student before analysis.

Lab Monitor Engagement Rating

Lab monitors were asked to rate student engagement in the course itself at the end of the experiment on a 5-point scale. Because of the lab monitors’ daily observations, they supplied a unique perspective for evaluating the extent to which the students were engaged with the course material. Ratings were based on lab monitors’ observations of student behavior throughout the experimental study and on their daily log of student behavior. Lab monitors were asked to consider three dimensions in their rating: the extent to which the student self-started in the course, the extent to which the student persisted in course material, and the extent to which the student refrained from non-course related activity.

Number of Hours Worked

We calculated the number of hours the students spent in the course. Students recorded starting and stopping times on timesheets, and the sum of the hours they spent working in the course was calculated from these timesheets. Lab monitors verified accuracy of timesheets on a weekly basis. Although we had asked students to spend approximately 50 hours to complete the course, they were free to spend less time as long as they worked consistently. Based on earlier research on motivation-building and caring/personal-investment interactions in the classroom—where self-efficacy, adaptive attributions, and caring teachers have tended to evoke increased levels of persistence on a task (Bandura, 1997; Weiner, 1986; Wentzel, 1998)—we expected that students who received enhanced messages from the teacher would remain engaged in the course material longer and that increases in effort and persistence would be reflected in the amount of time the students spent.

Final Grades

After the students completed the course, we recorded the final grades assigned by the instructor. Final grade was based on the average score of all work completed by the student up to that point. Because courses were being field-tested, course grades had no real consequence for the students, and their receipt was generally de–emphasized. We postulated, however, that students’ grades would give some indication of the level to which students engaged in the course. Because we asked students to simply “do their best,” we believed the students’ final grades would reflect the degree to which the student persevered in writing and revising to produce the best work possible.

Read more...

A PDF file of the full article is available. Contact: jrte@iste.org. Please specifiy Volume and Issue number and article name.

Copyright © 2001, ISTE (International Society for Technology in Education). All rights reserved.