Reviewing the Effects of Anonymity on Peer Assessment

• In Event: Peer Assessment Effects on Student Learning Outcomes With a Focus on Higher Education

Abstract

Objectives and Theoretical Framework
Peer assessment has many educational benefits for students such as gaining domain-specific skills, higher academic performance, and the development of assessment skills (van Zundert, Sluijsmans & van Merriënboer, 2010). Unsurprisingly, peer assessment is inherently a social activity requiring a mutual relationship of trust between assessor and assessee. It has been argued that the quality of peer assessment processes can be improved if peer assessment is anonymous (Peterson & Peterson, 2011). However, the impact of anonymity is not always positive (Yu, 2012). The aim of this empirical review is to establish the effects of anonymous peer assessment in different outcomes variables.
Data Sources
A literature search was carried out using PsycInfo and ERIC identifying 182 viable records. 27 publications remained after abstract analysis. After full text was read the final number of included publications was 14. The inclusion criteria were: (a) the study had anonymity as an independent variable; (b) included empirical results of anonymity interventions in relation to outcome variables; (c) at least one control group or a within-subject design with different interventional phases; (d) conducted in formal learning settings; (e) peer-reviewed; and (f) in English.
Data coding and analysis
Due to the small number of studies found and the large diversity in designs in terms of the type of control groups and the complexity of the variables surrounding anonymity (e.g., type of measured outcomes, tasks, directionality of anonymity) it was not feasible to conduct a meta-analysis. Therefore, a narrative review method was adopted using an in-depth analysis of the results and a count of the results directions (Cooper, 2010).
Results and Conclusions
The narrative review revealed that anonymous peer assessment seems to provide advantages for students’ perceptions about the learning value of peer assessment, delivering more critical peer feedback, increased self-perceived social effects, a slight tendency for more performance, especially in higher education and with less peer assessment aids. Conclusions are that: (a) when implementing anonymity in peer assessment the instructional context and goals need to be considered, (b) existent empirical research is still limited, and (c) future research should employ stronger and more complex research designs.
Significance
First, the empirical evidence on peer assessment anonymity suggests that the effect of anonymity on different peer assessment outcomes is far more complex than it has been assumed, especially in higher education. Second, research using anonymity seems to overlook the role of peer assessment constellations which become more relevant with university students. Third, many of the studies’ results are affected by the social and learning context that are not controlled for, as many of the available studies are quasi-experimental. Fourth, there is the classical tension between “laboratory” studies in which only one or two variables are manipulated where we would obtain more knowledge about how the individual variables modify the effects (e.g., just anonymity), and naturalistic studies in which peer assessment is implemented in the classroom with many variables playing a role, but with difficulties to disentangle different effects.

Authors

• Ernesto Panadero, Universidad de Deusto

• Maryam Alqassab, LMU Munich