Networking for Success dataset - project by University of Nottingham Malaysia

Institutional ethics approval was obtained (ANONYMISED FOR PEER REVIEW), and participation was based on voluntary informed consent. Data were anonymised prior to analysis and missing values per variable were small (1-15 cases only), addressed through pair-wise deletion. </p><p dir="ltr">Peer network data were collected using a web-based roster‑based survey design on Qualtrics, in which respondents were presented with a list of their cohort members (Bjorklund & Daly, 2021; Liou et al. 2016).

This approach, based on a similar instrument used to map pre-service teacher networks, reduces recall error and allows for the construction of bounded network representations. Following prior research, respondents reported peer ties across the following network functions:</p><ol><li><b>Advice‑seeking</b> (e.g. academic or professional advice)</li><li><b>Resource-sharing</b> (e.g. notes, materials, teaching resources)</li><li><b>Collaboration</b> (e.g. working together on academic tasks)</li><li><b>Emotional support</b> (e.g. peers turned to for encouragement or personal support)</li></ol><p dir="ltr">For each relational domain, respondents indicated the presence and frequency of ties along a scale of 1-4.

This instrument provided the raw data for analysis using Gephi (v0.10) (Bastian et al. 2009).</p><p dir="ltr">From the SNA survey, centrality measures were computed separately on Gephi for target network functions, exported in .csv format and treated as continuous variables in subsequent analyses on JASP software (v. 0.95.4) (weighted degree, betweenness centrality, eigenvector centrality). </p><p><br></p><p dir="ltr"><i>Academic success</i></p><p dir="ltr">Academic success was operationalised using within‑institution percentile rank, derived from institutional assessment records.

Because participating institutions employed different grading systems (e.g. percentage marks, cumulative grade point averages), percentile ranks were computed separately within each institution to represent students’ relative academic standing within their local assessment context. This approach preserves ordinal relationships while avoiding assumptions of metric equivalence across grading regimes (Brennan, 2006; Kolen & Brennan, 2014).</p><p><br></p><p dir="ltr"><i>Professional readiness</i></p><p dir="ltr">Perceived professional readiness to teach was measured using the Pre‑service Teacher Professional Experience (PTPE) scale, a validated self‑report instrument developed specifically for use with pre‑service teachers during or near the completion of their professional experience (Abraham et al., 2021).

The PTPE’s 36-item scale assesses pre‑service teachers’ perceptions of their preparedness to teach, readiness to engage with the profession, and teaching self‑efficacy, grounded in professional teaching standards and empirical research on teacher development. Items were rated on five‑point Likert scales.</p>