Author: Katerina Bohle Carbonell (Maastricht University)
Today at the European Social network conference was Learning day! Several consecutive sessions discussed the use of networks for learning research. This encompassed topics such as teacher education, teacher communities, knowledge transfer, acquiring knowledge and forgetting about relationships, mentoring, methods about learning etc. To summarize the sessions this network graph below shows the users (in blue) and the topics they talked about (in red). We have some interconnected presentations, researchers talking about similar topics, but we also have some isolate, outsiders or ‘Seiteneinsteiger’. The data for this graph is based on my humanly flawed memory about what was discussed during the session. Feel free to add users and keywords to the google sheet. I’ll update the picture using the code below.
What did we learn about teachers?
Bastian Later (Mainz University) showed us that teachers have different ties. Using the network flow model by Borgatti he demonstrated that the various relations teachers have are used for different purposes. Ties matter, but not all content flows through every tie. In connection to this Nienke Moolenar and Yvette Baggen (Utrecht University) highlight that teachers mental models are important. Especially, teachers who consider pedagogical content knowledge and strategy over theory as important skills teachers should have, are often thought out for advice on topics related to educational reform, in this case, specifically common core state standards. Following the line of teachers, Laura Thomas (Ghent University) demonstrates the importance of relationships to help young teachers not drop out of their profession. Finally, Lorena Ortega (University of Tübingen) and Zsófia Boda (ETH Zürich) talked about the teacher collaboration the exists to help at risk students to succeed in education. They demonstrated that there is a tendency to seek advice from members of the same department and that certain structural features explain collaboration. Two presentations analyzed online networks. Martin Rehm (Duisburg- Essen University) looked at what is going on inside these online communities. Together with colleagues he created a social brokerage index. This brokerage was defined by the in-and out degree of users, creating passive and active brokers. Passive brokers influence other users by having a central position in the network, while active broker seek to influence others by spreading information and thus having higher out-degree than in-degree. This index helps to explain why a certain person tweets a certain content. Following this line of looking at the content of tweets, Katerina Bohle Carbonell (Maastricht University) aimed to analyze the expertise distribution in a specific twitter teacher community (#acps). Transforming the twitter data into a user-topic network, she reported that the teacher community experienced low levels of specialization, and that users tweet about similar topics.
What did we learn about students?
Isabel Raabe (University of Oxford) looked at the STEM pipeline. The challenge for STEM profession is the sequential structure of milestones students need to conquer before being able to work in a STEM field. Along this path, individuals drop out, especially girls. One argument is that peers influence students subject choices and thus by having no friends who take STEM subjects students stop following these courses. She reported that indeed peers influence the choices students make about their subjects. Manuel Hopp (Friedrich-Alexander University of Erlangen-Nuremberg) looked at peer mentoring of students in STEM fields. He argued that the reputation of other peers in the mentoring community can influence a student to drop a STEM field. This is supported by the data. Finally, Vanina Torlo (University of Greenwich) indicated how faulty our memory of past relationships are. For each type of relationship a different churn profile was reported. Hence, all relationships change, but the degree to which they change over time differs.
What did we learn about us researchers?
We also learned something about us. Dominik Fröhlich (University of Vienna) presented how complicated good mixed method research can be. He is creating a social network map of how different methods are included in different studies. This will provide a guidelines for how to better conduct mixed method research.
This post has not been edited or peer-reviewed. Feel free to comment and suggest improvements!
R code to create graph:
df <- read.csv(“eu_sna.csv”, header=T, sep=”,”)
df <- df[,1:2]
names(df)[1]<- “user”
graph <- graph_from_data_frame(df)
V(graph)$name
V(graph)$type <- “topic”
ggraph(graph) +
geom_edge_link() +
geom_node_point(aes(colour = type))
ggsave(“eu_sna_2017_user_topic.png”)
Network For Learning 