Analysis Of Collaborative Conversation Characteristics Of Scratch Pair Programming In Primary School

In recent years,programming education has been widely recognized and is flourishing as one of the effective means of developing students’ overall skills,with a significant number of learners at a younger age and a large number of learners at the K-12 level.However,as a complex cognitive activity,programming requires a high level of personal ability,which has led to a high abandonment rate among beginners.Stakeholders have explored a wide range of learning styles,programming tools,and instructional content in an effort to lower the threshold for beginners to learn programming and optimize the effectiveness of programming instruction.Pair programming has been found to be one of the ways to enhance the effectiveness of programming instruction,significantly improving students’ willingness to continue learning programming and their performance,and optimizing their programming experience,especially for girls and struggling students.However,the individual characteristics of the collaborating partners can affect the collaborative activities and thus the effectiveness of pair programming.Therefore,in order to optimize the effectiveness of pair programming,it is important to identify the collaboration of students in pair programming activities.Thus,this study investigated the collaborative session characteristics of different programming ability combinations and gender combinations in pair programming with 14 fifth-grade students in an elementary school programming club in Jiangsu Province.The specific research process consisted of the following parts:First,we checked and combed domestic and international literature to understand the concept and current research status of pair programming,and clarify the factors affecting the effectiveness of pair programming in existing studies;understand theories related to collaborative problem solving and social role theories to provide a theoretical basis and relevant behavioral framework for the study,and develop a pre-test questionnaire with reference to the results of previous studies.Second,we measured and collected students’ programming abilities and collaborative intentions,and grouped students into groups;conducted teaching practice to familiarize students with pair programming while learning the basics;and had groups of students collaborate to create a game through pair programming in the last two lessons,and collected and transcribed and coded the dialogues generated during the period.Third,the data were analyzed using epistemic network analysis to characterize the collaborative sessions of groups with different programming levels in terms of overall session types,intra-group longitudinal development,and inter-group cross-sectional comparisons;to explore the collaborative session characteristics of different gender combinations in terms of overall session types,and to conduct in-depth analysis to explore the reasons from the perspectives of gender and role;and to explore the session transformation sequences of different types of combinations using lagged sequence analysis to explore the characteristics in terms of the structure and significant transformation sequences of both parties’ sessions,which complemented the results of epistemic network analysis.Fourth,based on the results of the data analysis,the following conclusions were drawn:(Ⅰ)Contribution-based discourse is the core of collaborative conversations regardless of the combination.(Ⅱ)The skilled-skilled group had obvious characteristics of session stages in the programming process,valued the implementation of coding solutions,had equal status of group members,communicated smoothly and at a deeper level,and had relatively flexible problem-solving styles.(Ⅲ)The skilled-novice group had more cognitive conflicts during programming,more reflections,and less irrelevant conversations.However,there was a clear primary and secondary relationship between group members’ status,and the conversations were dominated by question-and-answer activities but not deep enough,and the stage characteristics were not obvious.(Ⅳ)The novice-novice group generated more reflections and social conversations unrelated to activities during programming,and the group members had equal status and some depth of communication.However,limited by the lack of programming experience,programming activities progressed more slowly and often required help,and the type of conversation increased as the programming stage progressed.(V)The male-female group favors the planning and monitoring type,and has more efficient communication in programming;the female-female group favors the social and reflection and praise type,and has a better atmosphere in the group in programming.Finally,the following four pedagogical recommendations were made in response to the above findings:(I)The pair programming activities should make reasonable use of the characteristics of boys and girls combined with the characteristics of roles to match pairs.(Ⅱ)Teachers should teach programming knowledge while also focusing on developing students’ability to express and communicate their ideas.(Ⅲ)Provide support materials for students and groups with little programming experience before the programming activity begins.(Ⅳ)Emphasis on developing students’ self-and peer-supervision skills.