Research shows both benefits and challenges of online discussion as a collaborative learning activity. Online discussion is especially challenging for novice college students who have limited metacognitive skills as well as limited knowledge of the subject domain. With limited metacognitive skills, it can be challenging for novice students to engage in high-order cognitive activities and as a result, participate meaningfully in online discussions. In addition, in an online discussion with participants from different cultural backgrounds, the text-based nature of an online discussion medium and the cultural differences between participants may make it even more difficult to establish common ground and shared understanding than in an online discussion with participants from the same culture, especially if the participants do not have any or only have little experience with cross-cultural interaction. Therefore, the discussion board system in this study was designed with cognition-supported features (which combine a constraint-based discussion feature using a discourse structure with an automatically generated discourse map) and interactive intercultural elaboration features to address these issues. The study assessed the quantitative effects of these features on "knowledge construction", "critical thinking" and "intercultural sensitivity". In addition, the study also examined if prior knowledge and experience in the subject domain, prior experience in online discussions, or prior cross-cultural experience affected those results. Patterns of knowledge construction and critical thinking in four different types of online discussion boards were explored qualitatively by examining selected online discussion transcripts from eight weeks. Participants were 103 undergraduate students enrolled in an introductory-level psychology course at an Ivy League university in the northeastern United States who engaged in eight weekly online discussions in groups of eight students, each on a course-related topic. Students' knowledge construction in online discussions was examined using Gunawardena, Lowe, and Anderson (1997)'s Interaction Analysis Model that includes five phases of knowledge construction; students' critical thinking in online discussions was examined using Newman, Webb, and Cochrane (1995)'s critical thinking coding scheme with ten categories of critical thinking indicators. The results revealed that the cognition-supported features significantly helped learners in their knowledge construction and critical thinking in online discussion, although the effect was of greater magnitude for knowledge construction than for critical thinking. However, prior knowledge and experience in psychology and prior experience in online discussion did not influence the effectiveness of the cognition-supported features. Study results also indicated that the interactive intercultural elaboration features did not improve learners' intercultural sensitivity. Possible reasons for failing to show an effect of interactive intercultural elaboration features on students' intercultural sensitivity may have been the students' relatively high level of initial intercultural sensitivity and prior cross-cultural experience or students' imprecise use of the features. Results from descriptive analysis of four discussions in Phase 3 of the study to compare argumentation and problem solving discussion frameworks used as discourse structures revealed that in terms of knowledge construction, argumentation-based discussions had more evidence of low-level knowledge construction Phase I (Sharing and Comparing) and Phase II (Discovery of Dissonance), while problem solving-based discussions had more high-level knowledge construction Phase III (Negotiation and Co-construction) and Phase IV (Testing and Modification) evidence. This might have been due to the more structured nature of the argumentation framework. For both discussion frameworks, the group who used the cognition-supported features first tended to have more occurrences of higher phases than the group who only used these features towards the end of the semester. Study findings also revealed the possible influence of a discussion topic's characteristics on students' knowledge construction and critical thinking when comparing different discussion board types. In terms of knowledge construction, a discussion with a focus on finding and solving problems and with cognition-supported features had most occurrences of high-level phases and had fewest occurrences of low-level Phase I (Sharing and Comparing) during the knowledge construction process, followed by a discussion with a lot of details given initially but without any special features. On the contrary, a discussion with both sets of features but with a focus on sharing personal experience and from the group with less experience with cognition-supported features had the fewest occurrences of high-level phases but most occurrences of low-level Phase I. In addition, descriptive analysis of different discussion board types in terms of critical thinking showed that the discussions from the group more experienced with cognition-supported features had more higher level critical thinking indicators in the following categories: critical assessment, justification, and linking. The positive study results enable one to conclude that the constraint-based discussion and discourse map features should be used in online discussion activities in college classrooms and extended use of these cognition-supported features may be beneficial for students. (Abstract shortened by UMI.) [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com.bibliotheek.ehb.be/en-US/products/dissertations/individuals.shtml.]