Studying STEM Faculty Communities of Practice through Social Network Analysis

Spreading Teaching Innovations in Social Networks: the Bridging Role of Mentors

“…Using SNA, we examined the social network structures of individual CoPs. We found that the faculty CoPs in Programs I/II that began to use EBIP had strong ties in their teaching networks, whereas CoPs that did not use EBIP did not have these types of ties (Ma et al 2016;Ma et al 2018). These results suggest that strong ties were important for supporting the adoption of EBIP by CoPs; however, these studies did not explore how the CoPs learned about EBIP or how EBIP spread between CoPs.…”

“…These means and standard deviations were used to perform one-sample t-tests to compare the simulated networks with the mentorless network (Carsey and Harden 2013). We chose a conservative α-value of 0.001 for statistical significance to minimize the likelihood of Type I error for three reasons: our standard error of measurement was 1%, we perform multiple comparisons for significance with these permutation tests, and we have previously demonstrated the relationship of these cohesions metrics with the adoption of EBIP among CoPs (Ma et al 2018).…”

“…In previous papers, we documented the spread of EBIP between CoPs (Herman and Mena, 2015; and that COPs that adopted EBIP had statistically significantly higher density, higher connectedness, and lower centralization than CoPs that did not adopt EBIP (Ma et al 2016;Ma et al 2018). We determined which CoPs adopted EBIP by triangulating internal reviews provided by mentors, evaluations provided by an external evaluation team, and publication of education scholarship by the CoPs (Ma et al 2016;Ma et al 2018).…”

“…Network ties between two individuals can be treated as directed (information or relationship originates from a person and reaches another person) or undirected (information or relationships are treated as reciprocal). Because dyad reciprocity did not vary significantly between CoPs (Ma et al 2018), we treat all ties as undirected in our data.…”

“…Based on prior studies indicating that density, connectedness, and centralization (Borgatti et al 2002) were sufficient to distinguish between CoPs that adopted EBIP from those that did not (i.e., additional network measures such as breadth were nearly perfectly correlated with the chosen three measures; Ma et al 2018), we use these same metrics for the present study. Density is calculated as the actual number of ties in a network divided by the possible number of ties.…”