A review,analysis, and extension of peer-leader feedback agreement: Contrasting group aggregate agreement vs. self-other agreement using entity analytics and visualization

In reviewing peer-leader feedback within Multi-Source Feedback programs, the group aggregate agreement (GAA) method is contrasted with self-other agreement (SOA). Past research (Markham, Smith, et al., 2014) has demonstrated convergence problems with GAA for groups of peer raters. To evaluate dyadic convergence, we used the Benchmarks data to investigate two derailment factors (Building & Mending Relationships and Interpersonal Problems) for 4607 peers describing 1505 focal respondents. For high-agreement dyads, the r_total = ? 0.66** with 88% of the combined variance and covariance based on dyadic averages converging as whole units. Only 50% of all dyads demonstrated this type of high convergence. For low agreement dyads, the matching correlation (r_total = ? 0.56**) was almost exclusively a function of within-dyad divergence with only 4% stemming from between-dyad sources. Research implications for evaluating SOA under these agreement conditions are highlighted. Practitioner applications for using an entity-based visualization of dyads also are prototyped and discussed.