The Validy of Peer-Grading Assessment in MOOCs

Annotation on the topic of «The Validy of Peer-Grading Assessment in MOOCs» The peer-review is widely used in online courses. An important problem of this type of assessment is the limited amount of information about its validity and reliability. Purpose: measurement validity and identify distortions in the quest for mutual assessment used in MOOCs. Objectives: analysis of theoretical and empirical representations on the application of peer-assessment in MOOCs; measurement validity in the framework of the IRT; comparison of methods of research of mutual validity assessment in online courses. Data on the validity of the mutual evaluation studies are contradictory, which was a prerequisite for the research. Research problem: the existence of a mismatch between the active use of peer-assessment in the courses and the lack of sufficient data to confirm its validity, and as a consequence of this type of objective evaluation and a limited amount of data on the detection of distortions in the quest for mutual evaluation. Analysis revealed the assessment of experts, which are not correlated with the level of students' abilities and are biased. There were also identified experts who underestimated scores biased assessing the high level of students' abilities. Analysis of the criteria revealed that there is a need for greater rigor in assessing students using rubrics. We found evaluations, which may not reflect the real abilities of students and be objective. These estimates can be used for inclusion in the final assessment as they biased students underestimating score. Nevertheless, we have identified constraints, which have become a prerequisite for new research questions and implementation of the study with their account.