Nobuko Fujita

Statistical discourse analysis of online discussions: informal cognition, social metacognition and knowledge creation

To statistically model large data sets of knowledge processes during asynchronous, online forums, we must address analytic difficulties involving the whole data set (missing data, nested data and the tree structure of online messages), dependent variables (multiple, infrequent, discrete outcomes and similar adjacent messages), and explanatory variables (sequences, indirect effects, false positives, and robustness). Statistical discourse analysis (SDA) addresses all of these issues, as shown in an analysis of 1,330 asynchronous messages written and self-coded by 17 students during a 13-week online educational technology course. The results showed how attributes at multiple levels (individual and message) affected knowledge creation processes. Men were more likely than women to theorize. Asynchronous messages created a micro-sequence context; opinions and asking about purpose preceded new information; anecdotes, opinions, different opinions, elaborating ideas, and asking about purpose or information preceded theorizing. These results show how informal thinking precedes formal thinking and how social metacognition affects knowledge creation.

Socio-Dynamic Latent Semantic Learner Models

In this chapter we present a framework for learner modelling that combines latent semantic analysis and social network analysis of online discourse. The framework is supported by newly developed software, known as the Knowledge, Interaction and Social Student Modelling Explorer (KISSME), that employs highly interactive visualizations of interactions and semantic similarity among learners. Our goal is to develop, use and refine KISSME to generate and test predictive models of learner interactions to optimise learning.

Role taking and knowledge building in a blended university course

Role taking is an established approach for promoting social cognition. Playing a specific role within a group could lead students to exercise collective cognitive responsibility for collaborative knowledge building. Two studies explored the relationship of role taking to participation in a blended university course. Students participated in the same knowledge-building activity over three consecutive, five-week modules and enacted four roles designed in alignment with knowledge building pedagogy (Scardamalia and Bereiter 2010). In Study 1, 59 students were distributed into groups with two conditions: students who took a role in Module 2 and students who did not take a role, using Module 1 and 3 as pre and post tests. Results showed no differences in participation in Module 1, higher levels of writing and reading for role takers in Module 2, and this pattern was sustained in Module 3. Students with the Synthesizer role were the most active in terms of writing and the second most active for reading; students with the Social Tutor role were the most active for reading. In Study 2, 143 students were divided into groups with two conditions: students who took a role in Module 1 and students who did not take a role. Content analysis revealed that role takers tended to vary their contributions more than non-role takers by proposing more problems, synthesizing the discourse, reflecting on the process and organization of activity. They also assumed appropriate responsibilities for their role: the Skeptic prioritizes questioning of content, the Synthesizer emphasizes synthesizing of content, and the Social Tutor privileges maintaining of relationships. Implications of designing role taking to foster knowledge building in university blended courses are discussed.

Generating predictive models of learner community dynamics

In this paper we present a framework for learner modelling that combines latent semantic analysis and social network analysis of online discourse. The framework is supported by newly developed software, known as the Knowledge, Interaction, and Social Student Modelling Explorer (KISSME), that employs highly interactive visualizations of content-aware interactions among learners. Our goal is to develop, use and refine KISSME to generate and test predictive models of learner interactions to optimise learning.

Critical Reflections on Multivocal Analysis and Implications for Design-Based Research

This chapter presents critical reflections on the multivocal analyses presented in the preceding chapters in this volume by Teplovs and Fujita, Law and Wong, and Chiu on the asynchronous discussion data collected in an online graduate education course using Knowledge Forum. The multivocal analyses are discussed along five dimensions: theoretical assumptions, purpose of analysis, unit of analysis/unit of interaction, data representations, and manipulations on data representations. The diverse interpretations and findings of pivotal moments are explicated in light of broader dynamic group processes that support knowledge building in online graduate course contexts. The implications of multivocal analysis for design-based research are discussed.

Automating the analysis of collaborative discourse: identifying idea clusters

This poster explores CSCL practices relating to the use of a tool that employs information visualization techniques and large-scale text processing and analysis to complement qualitative analysis of collaborative discourse. Results from latent semantic analysis and qualitative analysis of online discussion transcripts are compared. Findings suggest that such tools that automate analyses of large text-based data sets can offer CSCL researchers a quantitative and unbiased way of identifying a subset of data to study in depth.

Online Graduate Education Course Using Knowledge Forum

Progressive discourse is a kind of collaborative discourse for inquiry in which participants share, question, and revise their ideas to deepen understanding and build knowledge. Although progressive discourse is central to knowledge building pedagogy, it is not known whether it is possible to detect its emergence in the participation patterns in asynchronous conferencing environments or what kinds of instructional interventions are most effective to support its development. To characterize episodes of discourse in which participants honor the commitments for progressive discourse and to refine designs of peer- and software-based scaffolding, the data used in this section was collected in the context of a study that examined student interactions on the asynchronous online discussion platform, Knowledge Forum®, in an online graduate educational technology course.

Online Learning Journals as an Instructional and Self-Assessment Tool for Epistemological Growth

This study looked at the instructional and assessment effects of using learning journals in three distance asynchronous computer conferencing courses (n=18, n=16, n=17). The instructor used a design-research methodology: each iteration of the course involved modifications to how learning journals were used based on analyses of the responses and results from the preceding course. Modifications included: a) use of orienting questions; b) question content, c) journal assessment and d) amount of scaffolding. Protocols were analyzed with a view to characterizing students’ epistemic cognition from two perspectives: belief mode (rationalist epistemology, self analysis, norms of inquiry to defend competing beliefs) and design mode (knowledge building epistemology, collective responsibility, norms of inquiry to support idea improvement and explanatory coherence). Changes in metacognitive reflection and learning journal activity were related to measures of learning. As a pedagogical tool, learning journals with directed questions (scaffolding) encouraged self-awareness of learning and epistemological reflection. Resume : La presente etude a examine les effets de l’utilisation de journaux d’apprentissage sur l’enseignement et l’evaluation dans trois cours a distance asynchrones en teleconference assistee par ordinateur (n = 18, n = 16, n = 17). L’instructeur a utilise une methodologie de recherche-conception : a chaque prestation du cours, des modifications etaient apportees a la maniere dont les journaux d’apprentissage etaient utilises en se basant sur l’analyse des reponses et les resultats obtenus lors de la prestation precedente. Les modifications concernaient : a) l’utilisation de questions d’orientation; b) le contenu des questions; c) l’evaluation du journal; d) la quantite d’echafaudage. Les protocoles ont ete analyses de maniere a caracteriser la cognition epistemique des etudiants a partir de deux points de vue : le mode « croyance » (epistemologie rationaliste, autoanalyse, normes d’enquete pour defendre les croyances concurrentes) et le mode « conception » (epistemologie de coelaboration des connaissances, responsabilite collective, normes d’enquete pour appuyer l’amelioration des idees et la coherence explicative). Les changements dans la reflexion metacognitive et l’activite des journaux d’apprentissage etaient lies a des mesures de l’apprentissage. En tant qu’outil pedagogique, les journaux d’apprentissage avec questions dirigees (echafaudage) encouragent la prise de conscience de l’apprentissage et la reflexion epistemologique.

Determining curricular coverage of student contributions to an online discourse environment through the use of latent semantic analysis and term clouds

This paper presents a new approach to mapping student contributions to curriculum guidelines through the use of Latent Semantic Analysis and information visualization techniques. A new information visualization technique - differential term clouds - is introduced as a means to make clear changes in semantic fields over time.