|
| Improvement of the Quantity and Quality of Undergraduates' Participation in Class Discussion |
| Sunday, May 30, 2010 |
| 4:30 PM–5:50 PM |
| Texas Ballroom Salon A (Grand Hyatt) |
| Area: EDC/TBA; Domain: Applied Behavior Analysis |
| Chair: Robert Lee Williams (University of Tennessee) |
| Discussant: Christopher Skinner (University of Tennessee) |
| Abstract: This symposium will provide an overview of different methods for assessing and improving both the balance of discussion across students and the quality of student comments in large college classes. The principal mode of assessing quantity of discussion was student self-recording of their class comments each day. Strengths and weaknesses of the self-assessment procedure will be addressed by examining its reliability of self-recording under credit and non-credit conditions for participation. The possible treatment effect of self-recording on participation also will be examined. In addition, the effects of a variety of credit conditions on the balance of discussion across students will be presented: (1) continuous daily credit for participation; (2) intermittent credit for participation across days within course units, with the credit days randomly determined at the end of each unit; and (3) intermittent credit across days within units, with the credit days randomly determined at the end of the course. In addition to examining the role of self-recording and credit on the distribution of discussion across students, we will propose a method for assessing the quality of discussion and identify predictors of discussion quality. |
| |
| The Effects of Self-Recording and Contingent Credit on Student Participation in Large College Classes |
| KATHERINE R. KROHN (University of Tennessee), Kathleen Briana Aspiranti (University of Tennessee) |
| Abstract: The study compared the effects of students’ (a) self-recording versus not self-recording class participation, (b) receiving credit versus no credit for participation, and (c) combining self-recording and credit on the quantity of class participation in three large sections of an undergraduate human-development course (initially 55 students per section). Percentage of students participating at different levels (e.g., non-participation, occasional participation, frequent participation, and dominant participation) constituted the major dependent variable. Ancillary dependent measures included relevance and type of class comments. In selected phases, students self-recorded their comments on specially designed record cards that included space for recording other credit-producing activities (e.g., homework completed). In other phases, students received a small amount of credit for making up to two comments in class discussion. Two different sets of observers recorded student participation and instructor behaviors that might affect participation. The research design included elements of both reversal and multiple-baseline designs. Credit decreased the percentage of both non-participants and dominant participants, thus balancing participation across students. Self-recorded participation strongly correlated with observer-recorded participation, but self-recording had minimal effect on participation. Neither treatment condition altered relevance or type of student comments. Instructor behaviors differed minimally across treatment condition and appeared not to inflate or diminish the effect of treatment conditions. |
| |
| The Effect of Random and Delayed Credit for Participation on the Consistency of Class Discussion |
| KATHLEEN BRIANA ASPIRANTI (University of Tennessee), Daniel F. McCleary (University of Tennessee), Katherine R. Krohn (University of Tennessee) |
| Abstract: This study was directed toward improving the consistency of student participation by thinning, randomizing, and delaying credit for student participation. The participants consisted of approximately 300 undergraduate students in a human-development course over a two-semester period. A benefit of not recording participation credit every day was to reduce instructor time in keeping track of student participation for grading purposes. Again, students used the specially designed record card to report comments and other credit-producing activities of the day. The major distinction between Presentations 1 and 2 was the inclusion of random schedules of credit and delayed credit in Presentation 2 as opposed to the continuous and immediate credit arrangement described in Presentation 1. The findings show that credit must be immediate when random schedules are used to achieve consistency of participation across time and students. The combination of randomized and delayed credit had little effect on the consistency of participation. |
| |
| Reliability and Predictive Validity of a New Qualitative Measure of Pariticpation in Class Discussion |
| LISA N. FOSTER (University of Tennessee), Daniel F. McCleary (University of Tennessee), Robert Lee Williams (University of Tennessee) |
| Abstract: Although a variety of studies have shown that contingent credit can increase the percentage of students participating in class discussion, there is little evidence that quality of discussion is affected concomitantly. In fact most studies that have examined ways to improve class discussion have not targeted quality, perhaps because of the difficulty in operationally defining and assessing quality. The current study developed an in-depth measure of quality that permitting grading students’ contribution to class discussion on a daily basis. The assessment procedure includes the type of questions and perspectives students voice in class, the types of responses students make to questions asked in class, and the types of examples students offer to illustrate or explain concepts in course materials. The study will determine which of the following dimensions are predicted by quality of student discussion: quantity of class discussion, attitudes toward participation, critical thinking, and test performance. If our newly developed measure of quality proves reliable and predictive, the next step will be to apply treatment procedures directly to quality to assess the possibility of a treatment effect. |
|
| |